Skip to main content

Full text of "Psychophysiological correlates of self-esteem"

See other formats


PSYCHOPHYSIOLOGICAL CORRELATES 
OF SELF-ESTEEM 



By 
GEORGE MICHAEL BEDINGER 



A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF 

THE UNIVERSITY OF FLORIDA 

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE 

DEGREE OF DOCTOR OF PHILOSOPHY 



UNIVERSITY OF FLORIDA 
1983 



To friends among the shade 

who selflessly gave so much for the uncaring, 

And who demanded fairness for others 

at great personal jeopardy, 

and whom I and we can never repay. 

Especially Phinny and Ed. 



ACKNOWLEDGEMENTS 

This dissertation is of course a beginning and not 
simply an end. But it is the epitaph of a very long and 
fruitful life goal. This has been an arduous search for 
insight into personality, a first step on a little marked 
trail. Although I have found that the search has often been 
lonely as one moves tree by tree, the forests are replete 
with those who will take a moment from their own pursuits. 
To all those who paused and shared their time and their 
ideas with me I express my appreciation. 

But certainly there were those who paused longer, and 
then journeyed a ways with me. Their ideas, support, and 
encouragement were invaluable and very special for me. 

Don Avila contributed the basic question I have so 
diligently researched. He first authored a grant proposal 
that succinctly stated the problems concerning research on 
self-esteem and how this research has failed to satisfac- 
torily address many of the empirical requirements of the 
modern scientific community. Further, he became a special 
friend, and my friends contribute the real joy and meaning 
in this life's too quick journey. 

Barry Guinaugh is a rock whose unerring judgement has 
served me well. Although we sometimes disagreed, time has 



amply demonstrated the wisdom of his advice. I reminisce 
often about the year we did our weekly behavior therapy 
group at the local mental health center. It was successful 
because he made it so. 

Bill Baxter supervised my mental hospital practicum. 
He piled a stack of appropriate literature in front of me 
and took the time to talk about it. Initially I observed 
from an unobtrusive corner as he conducted individual and 
group therapy with some of Florida's most difficult clients. 
But then most importantly, he gave me the freedom to learn. 
For nine months he turned me loose with individuals and 
groups of my own to fall down, get up, and learn to help 
people . 

Julian Keating pushed, set me up, and yet supported me 
unfaulteringly during several years of difficult labor. He 
is one of the few true philosophers , and one of the few who 
sees clearly the triviality when others are blinded by 
molehill mountains. In a maelstrom of conflict, he uner- 
ringly grasps the keystone. 

Harold Riker , a national figure in the study of geron- 
tology, took time from his busy schedule to listen and 
react to my ideas about the aging process. Of course many 
were first formed in his seminars. Speaking of which, he 
welcomed me into an extra seminar to refresh my memory and 
renew some old acquaintances. But the most appreciation is 
reserved for the innumerable times he took just to pause and 
interact as we happened together amidst our busy schedules. 



I have saved Bob Jester for the "last but not least" 
of my mentors. In our years as officemates I learned more 
statistics than in all my formal instruction put together. 
Somewhere we also became friends. The dozens of lunches 
with Don and him were welcome--even indispensable—islands 
of mental health in the ocean of insanity that is graduate 
school. 

Susan Angenendt , colleague, now spouse, has made 
beautiful music of my world professionally and personally. 
Family has to put up with so much and my parents, an uncle, 
and my wife did. 

Amidst the insanity and chaos of one's journey through 
time and despite and because of one's indispensable family 
of friends, one must every once in a while take time to say 
the hell with it all. This is my life, my journey, and 
I'm going to get on with it in my own good time. 



TABLE OF CONTENTS 

Page 

ACKNOWLEDGEMENTS iii 

LIST OF TABLES viii 

LIST OF FIGURES ix 

ABSTRACT x 

CHAPTER 

I INTRODUCTION 1 

Purpose of the Study 1 

Statement of the Problem 2 

Description of the Study 7 

II REVIEW OF LITERATURE 10 

Historical Review of the Psychophysiological 

Literature 11 

Independent Variables 19 

Well Being 19 

Depression 21 

Stress 23 

Dependent Variables 25 

Right Forebrain Electrical Activity ... 25 

Heart Rate 27 

Respiration Rate 29 

Perspiration Rate 30 

Blood Pressure 32 

Between Groups Factors 33 

Self-Esteem 33 

Aging 35 

Pilot Studies 38 

III METHODOLOGY 41 

Hypotheses and Corollaries 44 

Subjects 45 

Apparatus 47 

Instrumentation 47 

Procedure 48 

Laboratory Setting 48 

Instructions 49 

General Design Used in the Present Experiment. 51 



Page 

IV RESULTS 55 

Demographic Results 55 

Right Forebrain Electrical Activity 57 

Heart Rate 62 

Respiration Rate 64 

Perspiration Rate 67 

Blood Pressure 69 

Blood Pressure at Exit 74 

Reliability and Validity 79 

V DISCUSSION 85 

Dependent Variables 85 

Electroencephalogram 85 

Electrocardiogram 87 

Respiration Rate 88 

Perspiration Rate 89 

Blood Pressure 90 

Limitations of the Present Study 91 

Suggestions for Additional Research 93 

Implications for Education 94 

Implications for Theory 95 

Conclusion 98 

APPENDIX 

I SUBJECTS 101 

II INFORMED CONSENT 105 

III UNIVERSITY OF FLORIDA HUMAN SUBJECTS COM- 
MITTEE APPROVAL 10 7 

IV MEDICAL HISTORY AND SES QUESTIONNAIRE 109 

V INSTRUCTIONS TO SUBJECTS FOR MENTALLY IMAGING 
TREATMENT CONDITIONS 110 

VI RESULTS OF ANALYSIS OF VARIANCE PERFORMED ON 
SYSTOLIC BLOOD PRESSURE MEASURED SEPARATELY 
FROM PHYSIOGRAPH AND ITS RELATIONSHIP BETWEEN 

YOUNGER AND OLDER AGE GROUPS 114 

VII RESULTS OF ANALYSIS OF VARIANCE PERFORMED ON 
DIASTOLIC BLOOD PRESSURE MEASURED SEPARATELY 
FROM PHYSIOGRAPH AND ITS RELATIONSHIP BETWEEN 

YOUNGER AND OLDER AGE GROUPS 116 

BIBLIOGRAPHY 117 

BIOGRAPHICAL SKETCH 125 



LIST OF TABLES 



Table Page 

1 SUMMARY OF ANALYSIS OF VARIANCE PERFORMED ON 
RIGHT FOREBRAIN ELECTRICAL ACTIVITY MEASURED 

BY PHYSIOGRAPH (ELECTROENCEPHALOGRAPH) 58 

2 SUMMARY OF ANALYSIS OF VARIANCE PERFORMED ON 

HEART RATE MEASURED BY PHYSIOGRAPH 63 

3 SUMMARY OF ANALYSIS OF VARIANCE PERFORMED ON 
RESPIRATION RATE MEASURED BY PHYSIOGRAPH. ... 65 

4 SUMMARY OF ANALYSIS OF VARIANCE PERFORMED ON 
PERSPIRATION RATE MEASURED BY GALVANIC SKIN 
RESPONSE METER INDEPENDENTLY OF THE PHYSIO- 
GRAPH, BUT DURING EXPERIMENTAL TREATMENTS ... 68 

5 SUMMARY OF ANALYSIS OF VARIANCE PERFORMED ON 
SYSTOLIC BLOOD PRESSURE MEASURED BY 

PHYSIOGRAPH 70 

6 SUMMARY OF ANALYSIS OF VARIANCE PERFORMED ON 
SYSTOLIC BLOOD PRESSURE MEASURED SEPARATELY 
FROM PHYSIOGRAPH AND SEPARATELY FROM EXPERI- 
MENTAL TREATMENTS 75 

7 SUMMARY OF ANALYSIS OF VARIANCE PERFORMED ON 
DIASTOLIC BLOOD PRESSURE MEASURED SEPARATELY 
FROM PHYSIOGRAPH AND SEPARATELY FROM EXPERI- 
MENTAL TREATMENTS 76 

8 PSYCHOPHYSIOLOGICAL MEASUREMENTS AND THEIR 
RELATIONSHIP BETWEEN MENTALLY IMAGING DEPRES- 
SION AND GUILT (UNPLEASANT EVENTS), AND 
BETWEEN FIRST BP CUFF INFLATION AND HEARING 

DEATH STATEMENTS (STRESSFUL EVENTS) 81 



SUMMARY OF CORRELATION RESULTS : RELATIONSHIP 
BETWEEN SELF-ESTEEM (TSCS SCORE) AND PSYCHO- 
PHYSIOLOGICAL MEASUREMENT (DEPENDENT VARI- 
ABLES) BY TREATMENTS (INDEPENDENT VARIABLES) 
INCLUDING PSYCHOLOGICALLY SIMILAR MENTAL 
EVENTS (REPLICATION) 83 



LIST OF FIGURES 

Figure Page 

1. General design used in the present experiment . 53 

2. Design used to investigate differences between 
younger and older age groups 54 

3. Right forebrain electrical activity during 
experimental treatments and its relationship 

to self-esteem 61 

4. Systolic blood pressure during experimental 
treatments and its relationship to self- 
esteem 73 

5. Systolic and diastolic blood pressure measured 

at exit and their relationship to self-esteem . 77 

6. Theoretical model for construction of a 
self-concept instrument grounded in physiology. 99 



Abstract of Dissertation Presented to the Graduate 
Council of the University of Florida in Partial Fulfillment 
of the Requirements for the Degree of Doctor of Philosophy 

PSYCHOPHYSIOLOGICAL CORRELATES 
OF SELF-ESTEEM 

By 

George Michael Bedinger 

August 1983 

Chairman: Donald L. Avila 

Major Department: Foundations of Education 

The purpose of this study was to look for relationships 
between self-esteem and physiology. Self-theory has not made 
the contributions to education and counseling that it is 
capable of making because self-constructs have not been 
operationalized well enough to be adequate predictors of 
behavior . 

Seventy subjects (29 male, 41 female) were tested singly 
Right forebrain electrical activity (EEG) , systolic blood 
pressure (SBP) , heart rate, respiration rate, and perspira- 
tion rate were measured simultaneously while the subjects 
were mentally imaging well being and depression and under- 
going machine induced stress. 

Self-esteem as measured by the Tennessee Self-Concept 
Scale was significantly related to the EEG and SBP, but not 
to the other physiological indices. The EEG was positively 
related to self-esteem during the well being state and nega- 
tively correlated during stress. The SBP was negatively 
related to self-esteem during both depression and stress. 



Replication of the findings was accomplished by having 
selected subjects also mentally image guilt and listen to 
statements about death (instruction induced stress). Signi- 
ficant self-esteem correlations with the EEG and SBP ranged 
between .20 and .50. 

High and low self-esteem groups were compared using 
analyses of variance (ANOVA) and the low self-esteem group 
had higher EEG during stress and suffered higher SBP during 
both depression and stress. These differences probably ex- 
acerbated and did mask expected differences due to the aging 
process (ANOVA- -subject population included 53 younger and 
17 age sixty and older adults). 

After the debriefing, casual SBP was not related to 
self-esteem for the younger age group but was strongly 
related (.71) for the older age group. Casual diastolic 
blood pressure was related to self-esteem for all seventy 
subjects. These results and the ANOVAs suggest that self- 
esteem is related to the developmental process and accounts 
for at least half of the variance due to aging. 

The age related findings spoke to the meaningfulness 
of the relationship between self-esteem and physiology and 
suggested practical as well as theoretical significance. 
These were addressed and a theoretical model for construc- 
tion of a more adequate self-concept instrument was sug- 
gested. 



CHAPTER I 
INTRODUCTION 

Purpose of the Study 

The purpose of this study was to determine whether 
physiological correlates exist for self-esteem and, if so, 
to determine if those with a measured state of high or low 
self-esteem also significantly differ in measured physi- 
ology. For instance, people with low self-esteem may suffer 
higher blood pressure than those with high self-esteem. 

Since large numbers of subjects may produce significant 
relationships that have little or no practical effect, simul- 
taneous investigation was accomplished comparing younger adults 
and older adults. This comparison was intended to demonstrate 
practical or "real-world" strength or weakness. In other 
words, if physiological effects due to one's self-esteem are 
very meaningful, physiological differences expected between 
younger and older groups due to the aging processes should be 
masked or exacerbated by the physiological differences due to 
one's self-esteem state. For instance, younger people with 
low self-esteem may not have blood pressure that is signifi- 
cantly different from older people, and people with high self- 
esteem will have lower blood pressure than other groups, 
especially during depression and stress. Further, if this is 
so, the physiology of older people (e.g., blood pressure) 
should be related to self-esteem because older people have 
usually had a lifetime of having higher or lower self-esteem. 



-1- 



-2- 



Statement of the Problem 

An individual's self-esteem, a major factor in self- 
concept or self-image, is considered the single most impor- 
tant determinant of human behavior by many authorities 
(Combs, Richards, and Richards, 1976; Hamachek, 1978). 
Self-esteem influences every aspect of personal behavior, but 
especially such characteristics as mental health, levels of 
aspiration, learning, and delinquency (Combs, Avila, and Pur- 
key, 1978). In education, self-esteem may be the most impor- 
tant single factor in the success or failure of teaching and 
learning (Howsam, Corrigan, Denemark, and Nash, 1976). 

For many persons in the helping professions the solu- 
tion to many of our personal and social problems lies in a 
better understanding of the self. However, self- theory and 
the research based on it have not been as fruitful as they 
must be in order to gain more widespread acceptance and 
implementation. Avila (1980) said there are three major 
problems which have kept self-theory from making the sig- 
nificant contributions of which it is capable: 

1. Theoretical Constructs. The constructs and terms 
the position employs are too abstract. They are 
not anchored directly to empirical measurement 
(operationalized) and the theory does not always 
generate highly reliable prediction. 

2. Instrumentation. The instruments used to in- 
vestigate the basic research concerns of the 
theory are inadequate. Further, these instru- 
ments (mainly self -report, inference from a 
paper and pencil task, and observational re- 
port) are not sensitive enough for every re- 
liable prediction of real-world behavior. 



3. Application. Because of the poorly defined 
constructs and inadequate instrumentation, 
research of self-theory implementation in the 
classroom has generated mixed results or, 
worse, unwarranted conclusions that have not 
been substantiated upon replication. (p. 2) 

Avila (1980) then said these weaknesses may appear to 
be so serious as to suggest that self-theory should be aban- 
doned. The understanding of self-esteem, a key and integral 
component of the theory, however, is so critical according 
to Avila, that few individuals can be found who are willing 
to propose such a course. Even Kenneth W. Spence , one of 
the most ardent behaviorists and a rigorous experimental 
psychologist, has defended self-theory: 

That this field approach to the problems of 
psychology has been fruitful and valuable is 
amply supported by the experimental contribu- 
tions it has made. . . . Furthermore, the 
phenomenological approach has its advantages, 
particularly in the complex field of social 
behavior of the human adult. (Spence, 1963, 
p. 170) 

Ruth Wylie (1974a), probably the severest critic of the 
theory, has added that even though the process of giving 
the position a sounder scientific basis is going to involve 
many long and arduous tasks, "I believe these tasks can be 
accomplished and that their probable contribution to the 
science of personality is worth the strenuous effort 
required" (p. 316). 

Thus there seemed to be two major reasons for 
the present state of affairs in self-theory. The 
first has to do with the position from which the 
theory approaches the study of the human being. 



-4- 



Self-theory focuses on the internal aspects of the indi- 
vidual, being concerned with what goes on inside the orga- 
nism. Constructs based on internal conditions which have 
not yet been open to direct observation are extremely dif- 
ficult to translate into manipulable and controllable com- 
ponents. In order to render such constructs open to 
manipulation they must be tied to some type of operational 
procedures and translated into intervening variables in 
much the same way that behaviorists in psychology have done 
with considerable success for more easily observable 
phenomena . 

The second reason is related to the first. Because the 
basic constructs have had an inadequate empirical base, and 
then have not been sufficiently accessible to experimental 
manipulation, satisfactory tools of measurement could not 
be developed. 

Avila (1980) says this current dilemma for self- theory 
can be resolved with a technological and procedural model 
which can operationalize the position's constructs and that 
these are available but have not yet been investigated. He 
says adequate physiological bases have not been demonstrated, 
but must be there, given that the theory has real-world 
viability. Further, technological advances in the measure- 
ment of physiology now make sophisticated comparisons more 
easily accomplished. 

Recent years have witnessed a tremendous increase in 
interest and research in psychophysiological processes and 



their relationship to observable behavior. Studies along 
these lines have proliferated (Benson, 1975; Schwartz and 
Beatty, 1977) and a whole new discipline based on psycho- 
physiological measurement, biofeedback theory, has emerged. 
Most of this research, however, has been clinical and 
applied in nature and its tenets are probably premature 
since the results so far produced are often contradictory. 
The main purpose of most of these kinds of research has 
been to discover the consequences of clinical treatment. 
Very little basic research has been done and researchers in 
the area are confronting the same problems faced by self- 
theorists relating to poorly defined constructs. 

Of importance to the present dissertation is that 
psychophysiology has produced a technology which lends 
itself well to looking at internal aspects of the indi- 
vidual and then to the operationalization of self-theory 
constructs. By employing this technology it now seems 
possible that self-theorists will be able to find the em- 
pirical basis that is needed for the adequate development 
of the theory and its constructs such as self-concept and 
self-esteem. Then it will be possible to construct more 
accurate instrumentation. 

The psychophysiological model and psychophysiological 
techniques lend themselves well to the problems related to 
a holistic theory beginning with and concentrating on in- 
ternal variables. Self -theory is concerned with the nature 
of a subject's subjective experience, the internal processes 



-6- 



occurring and how these influence behavior. Gale (1973) 

states that this is precisely what the psychophysiological 

model allows one to study: 

Three different aspects of the person may be 
studied concurrently, along with a common time 
scale: performance or behavior, verbal report 
or subjective experience, and physiological 
state. The capacity to study events in three 
universes at once is the hallmark of the psycho- 
physiologist. His job is to construct units and 
scales of measurement which enable him to make 
sense of what is occurring concomitantly in all 
three universes: (i) physiology, or what is 
going on in the nervous system (e.g., as measured 
by variation in heart rate, electrodermal acti- 
vity, respiration, muscle activity and the EEG) , 
(ii) what the subject is observed to be doing 
(reaction time, learning, social behavior, 
activity level, and so on) and (iii) subjective 
report of experience (what is thought, felt or 
imagined). (Gale, 1973, p. 215) 

For this researcher the psychophysiological model 
gives a highly practical frame of reference for the study 
of self-theory and biomonitoring equipment provides exactly 
the kinds of tools the self- theorist needs in order to de- 
velop (1) an empirical basis for the constructs of the 
theory, (2) instruments that are sensitive enough to measure 
changes in these operationally defined constructs, and 
consequently, (3) strategies for achieving the goals of the 
theory. Lazarus (1977) sees the same potential when he 
states -. 

In the biofeedback laboratory or clinic, the per- 
son is given information about the activities of 
his visceral systems and asked to regulate these. 
How he/she does this, what works and what does 
not work, the limits of the effects in magnitude 
of control, over time--all such information is 
capable of contributing something of immeasurable 
value to our knowledge. Research using biofeed- 
back procedures could help us discover much more 



-7- 



than we now know about the psychological mechanisms 
of self-regulation, particularly the intrapsychic 
ones. (p . 85 ) 

Lazarus goes on to point out the feasibility and the 

necessity of the kind of research conceived in the present 

dissertation: 

If the problem is approached only in a parochial 
way, or as merely a gimmick limited to the bio- 
feedback laboratory, then we are likely to ad- 
vance little in spite of the evident potential. 
Biofeedback research will go much farther and 
rapidly become an integral part of psychology 
if it is seen and approached within the larger 
context in which it belongs. (Lazarus, 1977, 
p. 85) 

Thus the technology of psychophysiology is available 
to investigate whether physiological correlates exist for 
self-theory constructs. If self-theory constructs can be 
grounded in physiology then the theory will become more 
predictive of human behavior. Preliminary research looking 
for relationships between self-esteem and human physiology 
is then a giant first step towards solving self-theory con- 
struct operationalization problems. 

Description of the Study 

A self-theory construct, self-esteem, delineated by a 
paper and pencil task that is currently accepted as the 
most reliable and valid instrument, was utilized to rank order 
seventy subjects by self-esteem. Division of this order into 
high, middle, and low self-esteem groups permitted between 
group comparisons by measured physiology. 



Concurrent differences in physiology were produced by 
having the subjects mentally image separate psychological 
events, such as well being and depression. Behaviorally 
produced stress was also physiologically measured. 

With these measurements, correlations between self- 
esteem and physiology (such as blood pressure and electro- 
encephalogram) while the subjects were in differing psycho- 
logical states (well being, depression, and stress) was 
calculated. Further, the measured physiology was averaged 
for each self-esteem group and comparisons were made between 
high and low self-esteem groups during the different psycho- 
logical states. These calculations showed correlations or 
lack thereof between self-esteem and physiology during dif- 
ferent psychological states. They also showed self-esteem 
group differences or lack of differences during separate 
psychological states (e.g., high self-esteem subjects having 
higher or lower relative brain activity during well being 
or stress than do low self-esteem subjects). 

Finally, younger and older age groups were delineated. 
This grouping allowed comparisons between age groups to 
determine physiological similarities and differences (such 
as the younger age group having blood pressure as high as 
the older age group during the experiment). This last group- 
ing then made possible inferences concerning the "real- 
world" meaningfulness of the effects of self-esteem on 
physiology. In order to make the age group comparisons, 
the subject population, relative to the general population, 



had a disproportionate number of persons over the age of 
sixty. (One-quarter of the subject population was in the 
older group . ) 

After the experiment was over, blood pressure was taken 
by the conventional method and the results were given to the 
subjects as a "reward" for "having been such a good subject. 1 
These data, collected after the experiment, were utilized to 
verify expected a priori differences between the age groups. 
In other words, if the older group's mean blood pressure was 
higher than the younger group's mean after they all thought 
the experiment was over, the inference can be made that the 
older group had higher normal blood pressure. 



CHAPTER II 
REVIEW OF LITERATURE 



Human emotions, including personality variables such 
as "ego," have long been thought to be major factors in 
some physical disorders (Cannon, 1928). Researchers over 
the decades have investigated physiological correlates of 
many emotions (Sternbach, 1966). Further, the journal 
Psychophysiology is fully accepted by the American Psycho- 
logical Association, an event attesting to the present day 
sophistication of the field. 

Investigation in psychophysiology has often entailed 
simultaneous dependent measures of one or more emotions. 
The present design used both simultaneous measurement and 
delayed measurement and is most easily understood when the 
psychophysiological measures (e.g., blood pressure) are 
presented singly to the reader. Further, investigation of 
particular psychological states has taken place over decades 
and strict chronological reporting is hard to follow if the 
reader is led from one state to another and back to the 
first again. For these reasons strict adherence to chrono- 
logical reporting in this review was not done. 

First, a brief overview of psychophysiological history 
is presented. Then, additional historical and current 



10 



11- 



literature related to this design's independent variables 
(experimentally induced states of consciousness: well 
being, depression, and stress ) are presented. Some redundant 
presentation is then unavoidable but was held to a minimum. 
Next, literature related to the dependent variables (physio- 
logical measures) is presented. Again, some redundancy, 
primarily attributable to simultaneous measurement, was 
unavoidable. Finally, self-esteem literature followed by 
pertinent gerontological literature is included. 



Historical Review of the Psychophysiological 
Literature 



Cacioppo and Petty, writing in the May, 1981, issue of 
the American Psychologist , said that "the earliest writings 
to address the relationships between psychological and 
physiological phenomena are probably those of the ancient 
Greeks (e.g., about 500 B.C. in Plato's Theatetus j cf. 
McGuigan, 1978, Chap. 2; Mesulum & Perry, 1972)" (p. 441). 
But empirical research is relatively recent, begun about 
100 years ago--Cacioppo and Petty reference Angel and 
Thompson, 1899, and Sechenov, 1878-1947, for early reviews. 
The May, 1981, American Psychologist also contains a brief 
history of the field of psychophysiology , a definition of 
the field delineating boundaries between it and other fields 
in psychology, and an excellent discussion of the methods 
and applications of the field to psychological research. 



-12- 



Sir William Osier wrote in 1897 about the relationship 
of stress to arterial degeneration. He said he believes 
psychological variables influence physiology: "In the worry 
and strain of modern life arterial degeneration is not only 
very common, but develops often at a relatively young age. 
For this I believe that the high pressure at which men live 
and the habit of working the machine to its maximum capacity 
are responsible rather than excesses in eating or drinking" 
(p. 153-154). 

However, it was not for another generation that pro- 
cedures acceptable to today's scientific community began to 
be implemented. Baselines, the physiological measurements 
taken prior to an intervention, were taken and experimental 
manipulation such as learning and dreaming states were 
accomplished. These mental states resulted in increases 
in physiological responding over the baselines. This pro- 
cedure was detailed by Cacioppo and Petty (1981), who cited 
Clites (1936), Freeman (1930), and Golla (1921). Cacioppo 
and Petty then wrote about the 1950s, "As a result, several 
elaborate theories regarding task performance and arousal 
(in one of its many forms) were developed" (p. 441). They 
cited Duffy (1957), Lindsley (1952), and Malmo (1957). 

Returning to the 1930s, the Menningers hypothesized 
in 1936 that repressed, aggressive tendencies lead to heart 
disease. Mittelmann and Wolff (1939) showed that emotional 
conversation lowered hand temperature in those with Ray- 
naud's disease (cold hands). 



-13- 



The 1940s saw Dunbar (1943) describe coronary patients 
as hard driving, goal oriented people who are "workaholics" 
(in today's vernacular). At about the same time Bettelheim 
(1943) observed a different kind of interplay between 
psychological state (seemingly acute instead of chronic) 
and physiology. He described the "Muselmaner" (walking 
corpses) in Nazi concentration camps who, "because of their 
extreme sense of hopelessness, developed symptoms of apathy 
and withdrawal that many times resulted in death due to no 
known organic cause" (p. 417). Richter (1957) and Seligman 
(1975), researching "learned helplessness," also documented 
instances of sudden, unexplained death in animals that had 
no control over a stressful environment. 

Returning to the literature on coronary prone indi- 
viduals, Kemple (1945) said such individuals manifest "a 
persistent pattern of aggressiveness and drive to domi- 
nate. . . . They are usually very ambitious and strive 
compulsively to achieve goals incorporating power and 
prestige" (p. 87 ) . 

Probably the first research to directly observe changes 
in internal physiology that were produced by manipulating 
psychological states was done by Wolf and Wolff in 1947. 
They observed a subject with a gastric fistula (an opening 
in the stomach wall through which they could observe the 
stomach lining). They reported that stomach movement, gastric 
secretion, and dilation of blood vessels increased during 
anger and decreased during fear. Two emotional states 



IV 



(anger and fear) seemed to elicit the only kinds of patterns 
found. Although a great variety of emotions were manipu- 
lated, all others seemed to elicit more or less of the same 
two general patterns. 

However, seemingly confounding research was being pub- 
lished about the same time. Shaffer (1947) interviewed World 
War II combat pilots and found that most (80%) were easily 
irritated or angry while experiencing fear. This suggests 
that the physiologically different emotions found by Wolf 
and Wolff are involved in mutual "feedback loops" where one 
emotion affects another which in turn affects the former and 
a cycle was continued. 

Further evidence supporting Wolf and Wolff was found by 
Ax in 1953. Ax connected subjects to a polygraph under the 
pretext of recording their physiological responses while the 
subjects were relaxing. During the recording session, he 
provoked his subjects to intense anger by having the polygraph 
operator rudely insult them and to intense fear by leading 
them to believe that the polygraph was short-circuited and 
might electrocute them. Ax said respiration rate and per- 
spiration increased more during fear and blood pressure in- 
creased more during anger. However, these findings showed 
quantitative differences in the same direction and, although 
they certainly imply psychophysiological differences for 
different emotional states (or at least that different in- 
structions produce different results), this may not turn out 
to be Ax's most important contribution. Perhaps his most 



■15- 



important finding is one not often reported in the literature 
of psychophysiology although all report Ax's finding that 
anger and fear are psychophysiologically different emotions. 
Ax "found greater between-subject than within- subject vari- 
ance in physiological reactions which suggests that people 
may have distinctive modes of responding physiologically" 
(Cacioppo and Petty, 1981, p. 443). 

Lacey and Lacey showed in 1958 (as reported by Cacioppo 
and Petty, 1981) that 

there are multiple psychologically important fac- 
tors that influence the various physiological re- 
sponses at each moment in time. Two principles 
identified by the Laceys are individual response 
stereotypy and stimulus response stereotypy. 
Individual response stereotypy refers to the 
tendency for the same individual to display the 
same profile of physiological responses regard- 
less of the situation or stimulus, whereas 
stimulus response stereotypy refers to the 
tendency for a situation or stimulus to elicit 
a common pattern or profile of responses from 
people in general. (p. 443) 

Graham (1955) reported that significant forearm tem- 
perature decreases occurred during states of anger, stress, 
and depression. Graham, Stern, and Winokur (1958) found 
that hypnotically suggesting an attitude associated with 
hives resulted in raising hand temperature and that sug- 
gesting Raynaud's disease decreased hand temperature. 

Friedman and Rosenman (1960), after several years of 
study of occupational "risk factors," found that coronary 
artery disease was seven times more prevalent in a personality 
structure they had labeled "Type A" than in those they had 
labeled "Type B." Extensive interviews were used to 



■16- 



differentiate who was Type A, or coronary disease prone, 
and who was Type B. 

Friedman and Rosenman tried to identify Type A and 
Type B by using a polygraph, which measured respiration, 
body movements, and hand clenching, while subjects listened 
to tape recordings designed to elicit Type A behavior. This 
technique often misclassif ied more coronary patients as 
Type B rather than Type A. However, newer techniques such 
as voice analysis appear to be quite promising for assessing 
the A/B behavior dimension (Schucker and Jacobs, 1977). 

T.G. Burish (1981), writing in the Encyclopedia of 
Clinical Assessment (Volume 1), said the crucial link between 
Type A behavior and coronary heart disease may be the develop- 
ment of damaging psychophysioendrocrinological conditions. 
He cited studies showing significantly greater beta-lipo- 
protein concentrations, faster blood-clotting times, greater 
heart-rate variability, higher serum cholesterol and serum 
lipid levels, increased epinephrine and norepinephrine 
levels, and autopsy findings revealing more atherosclerosis 
and coronary occlusion in Type A individuals than Type B 
individuals. These studies seem overwhelmingly conclusive. 

Burish (1981) further said that there are two psycho- 
logical assumptions involved in the above conclusions. The 
first is that stress can lead to damaging physioendocri- 
nological consequences and the second is that Type A indi- 
viduals live under increased stress because of their Type A 
behavior (for instance, "workaholicism") . Burish then 
quoted Glass (1977) who hypothesized a definitive link 



-17- 



between Type A behavior and learned helplessness, a depres- 
sive reaction. 

The 1960s saw a marked increase in the level of sophis- 
tication of psychophysiological studies. Stern et al . 
(1961) identified a "startle effect" in psychophysiological 
measurement that can confound data. 

Schachter and Singer (1962) showed that there is an 
interaction between physiological arousal and emotionally 
toned cognitions. Nisbett and Schachter (1966) demonstrated 
that genuine emotions can sometimes be suppressed when 
people are given alternative explanations for their arousal. 
In other words, how one perceives the emotional situation 
will affect that emotion. 

"Real" arousal is not necessary according to Valins 
(1966), only the self-perception that one is aroused. Emo- 
tional disorders such as shyness, stuttering, and even 
impotence are characterized by a "feedback" cycle according 
to Davison and Valins (1972) . Self-perception then is an 
important variable, in addition to perception of the emo- 
tional situation. 

Lang et al. (1973) showed that there are reliable dif- 
ferences between the autonomic patterns shown in fear, anger, 
the startle response, hunger, and pain. They said that 
differences among the more subtle emotions have not been 
demonstrated consistently. They also found further evidence 
that Ax (1953) was correct in reporting consistency of 
responding within individuals. Lang et al. (1973) said 



• II 



some people respond with an all-out autonomic arousal to 
every emotion-producing situation, others seem to respond 
very little in any situation, and still others respond 
greatly to some kinds of situations and only slightly in 
others. Scientific classifications simply substantiate what 
people have known for centuries-- that there are definite 
emotional types of individuals. Perhaps if a consistent 
kind of responding can be delineated for groups of people, 
then a "personality label" can be ascertained (or invented) 
and highly accurate prediction of emotion will be possible 
for people within a particular group . 

Some of the more subtle responses Lang was referring to 
in 1973 have since been identified. "Significant differ- 
ences in hormone responses to such stress-producing factors 
as physical exertion, fasting, and exposure to intense heat 
and cold have been found" (Mason, 1975, p. 413; Mason et al . , 
1976). Certainly there should be even more subtle specific 
hormone patterns that will eventually be found for specific 
emotions (Lazarus et al., 1980). Dienstbier (1979) said 
that "the more important role will eventually be assigned 
to physiological patterns as a means of identifying specific 
emotional states and differentiating them from nonemotional 
stimuli" (p . 10) . 

Well being, depression, and stress were the specific 
emotional states investigated by many researchers. However, 
none investigated self-esteem and emotional states. 



-19- 

Independent Variables 
Well Being 

Clynes (1974) showed links between imagined emotional 
states and simple observable behavior. During an imagined 
state of love subjects pushed a key down slowly and away 
from themselves. While imaging anger the subjects pushed 
they key down straight and fast. A no emotional control task 
gave an intermediate response: The key was pushed straight 
down slowly. The physiological responses to love and anger 
were consistent across cultures. 

Light (1981) reported that young, healthy males who 
are very reactive people during stress are indistinguish- 
able physiologically (heart rate and blood pressure) from 
the less reactive persons when relaxed and feeling good 
(well being) . 

Relaxation and well being seem to be states primarily 
produced by the parasympathetic nervous system (PNS) according 
to Krech et al. (1974). The neurotransmitter is acetyl- 
choline in the PNS as it is in the somatic system. These 
function as part of the autonomic nervous system (ANS) . 

The PNS slows the system (e.g., heart rate) down while 
the sympathetic nervous system (SNS) speeds it up. Sur- 
prisingly, both can produce feelings of well being. The 
SNS activates both the adrenal medulla and the adrenal cor- 
tex while the SNS itself is activated in the brain by 
norepinephrine. The adrenal medulla produces epinephrine 



-20- 



and norepinephrine which "speed" the physiological system 
up. Many people experience well being, but usually not 
relaxation, when the adrenal medulla is stimulated. Addi- 
tionally, norepinephrine has been implicated in mania, a 
well being state (Krech et al., 1974). 

Further, well being can be produced by the SNS when 
the anterior pituitary hormone (adrenocorticotropic hormone 
or ACTH) stimulates the adrenal cortex to produce adrenal 
steroids. Although many people experience well being, many 
experience irritability and others no effect when given 
ACTH or adrenal steroids (Krech et al., 1974). 

Well being then, a psychological state, can be produced 
by either the PNS or the SNS but is usually associated with 
the PNS, while the SNS is associated with the stress 
response. Other neurotransmitters can produce well being 
(serotonin and the enkephalins) but their activating hor- 
mones or psychological events are not yet known (Kretch et 
al. , 1974). 

Levenson and Ditto (1981) investigated sixteen kinds 
of instructions to individuals to elicit heart rate (HR) 
changes. Only two were significantly related to performance, 
"Make yourself feel relaxed" (r = -.29 with HR) and "Think 
about something peaceful" (r = -.30) produced heart rate 
decreases . 

Levinson and Ditto (1981) also found that personality 
variables (locus of control, state and trait anxiety) were 



-21- 



not related to the ability to control heart rate increase 
or decrease. 

Cacioppo and Petty (1981) said that in 1976 Swartz, 
Fair, Salt, Mandel, and Klerman replicated Darwin's sug- 
gestion that distinctive facial expressions are linked to 
different emotions. "Nondepressed subjects displayed pat- 
terns similar to those produced by the depressed subjects, 
but the pattern accompanying pleasant imagery was accentu- 
ated and the pattern accompanying unpleasant imagery 
attenuated in normal subjects" (p. 443). 

Depression 

The literature on depression is voluminous. One of the 
best historical reviews is presented by Fabry (1981) in the 
Encyclopedia of Clinical Assessment , Volume II. 

Directly relating to the present study, Fabry (1981) 
said that Bibring in his 1953 study of depression examined 
the relationship between anger or hostility and depression. 
Bibring showed that when anger is turned inward or remains 
unexpressed, it retards behavior. When it is turned out- 
ward, it is manifested through agitated behavior. Fabry 
(1981) says Beck (1961) proposed that a negative view of the 
self, the world, and the future, along with self-blame and 
criticism is the primary element in depression. As pre- 
viously referenced, Seligman in 1975 said that depression 
results from "learned helplessness": The organism "gives 



•22- 



up" when confronted with an uncontrollable environment 
(hopelessness) . 

• Depression is also endocrinologically related. Reduced 
levels of norepinephrine (a neurotransmitter) often charac- 
terize depression and drugs that deplete norepinephrine 
produce depression (Schildkraut and Kety, 1967; Schildkraut 
and Freyhan, 1972). Dienstbier (1979) found the inducation 
of anger stimulated production of norepinephrine and counter- 
acted depression. 

Schuyler (1974) said that "normal depressive reactions 
become neurotic when the person shifts his attention from 
the significant other to the self" (p. 36). Fabry (1981) 
calls this reactive (situational) depression and labels more 
chronic depression as endogenous depression. Cammer (1972), 
writing about chronic depression, said "This type of de- 
pression has also been associated with postpartum depres- 
sion, aging, toxif ication, infectious diseases, glandular 
disorders, severe injuries, surgery or changes in body 
structure" (p. 14). "However, it has been most closely 
related to involutional melancholia at menopause" or aging, 
according to Fabry (p. 591). He said, "In general the in- 
volutional is characterized by a rigid, perf ectionistic 
life style" (p. 21) . 

Cacioppo and Petty (1981) found that nondepressed sub- 
jects displayed facial electromyographical (EMG) activity 
that was similar to prior EMG that they had displayed while 
they were imaging a pleasant experience when they were 



-23- 



thinking of their typical day. Moreover, depressed subjects 
showed their "depressive facial EMG" when they were asked 
to image their typical day. Schwartz et al . (1978) found 
that those most likely to improve clinically had resting 
facial EMG levels higher than those who showed little 
clinical improvement. Perhaps this implies those most likely 
to show clinical improvement are in an "anger" phase and 
have not yet reached an extreme learned helplessness (apa- 
thetic) phase. 

Krech et al. (1974) said that epinephrine (SNS) secre- 
tion increases when people are angry at themselves in a 
stressful situation. However, when people are angry at 
others or the situation they seem to have an increase in 
norepinephrine (also SNS). Thus anger, as a psychological 
state, may itself be as complex as depression. 

Stress 

Those suffering from chronic stress have been described 
in the historical literature as having "state anxiety" 
(Marinelli, 1981). Spielberger (1975) described state 
anxiety as "subjective, consciously perceived feelings of 
tension, apprehension, and nervousness accompanied by or 
associated with activation of the autonomic nervous system" 
(p. 137). Marinelli (1981) said state anxiety is expected 
to fluctuate in intensity over time. 



-24- 



Marinelli (1981) in his historical review of anxiety in 

the Encyclopedia of Clinical Assessment said: 

Freud (1926, 1936) led anxiety into the twentieth 
century by giving it a central position in his 
theory of personality. Freud's focus was on 
anxiety as a global motivational force rather 
than on the experience of anxiety. In building 
his theory, Freud depended primarily on 
hypothetico-deductive reasoning based on clini- 
cal observation. Learning-oriented theorists 
(Hull, 1921, 1943, 1952; Mowrer, 1939, 1950) 
made the first significant movements in using 
experimental methods for the theoretical study 
of anxiety. Important postulates in their con- 
ceptions are that (1) anxiety is, to a large 
extent, learned behavior; (2) it motivates 
trial-and-error behavior; and (3) its reduction 
reinforces the learning of new habits. These 
points are considered the touchstone of the 
drive conceptions of anxiety proposed by Dollard 
and Miller (1950) and Spence (1956, 1960). 
(p. 560) 

Hilgard et al. (1979) listed the effect of stress on 

the sympathetic nervous system (SNS) , a part of the autonomic 

nervous system (ANS) : 

1 Blood pressure and heart rate increase. 

2 Respiration becomes more rapid. 

3 The pupils of the eyes dilate. 

4 Perspiration increases, while secretion of 
saliva and mucus decreases. 

5 Blood-sugar level increases to provide more 
energy . 

6 The blood is able to clot more quickly in 
case of wounds. 

7 Motility of the gastrointestinal tract de- 
creases; blood is diverted from the stomach 
and intestines and sent to the brain and 
skeletal muscles. 

8 The hairs on the skin become erect, causing 
"goose pimples." (p. 330) 



-25- 



Interestingly, anger (also SNS stimulating) was almost 
the same as stress except for Hilgard's number seven: anger 
increased gastro-intestinal activity (Wolf and Wolff, 1947; 
Dienstbier, 1979). 

Gersten et al. (1974) showed where both positive and 
negative change produced symptoms of anxiety but only nega- 
tively weighted change correlated with their stress symptoms. 
Gersten et al. (1974) then imply a self-perception variable 
related to the physiological changes that they found. 

Cronbach and Snow (1977) said that self-esteem is simply 
the other side of the coin with anxiety. They posited a 
single morale factor, "constructive motivation." Corno et 
al. (1981) said "there might be more value in characterizing 
students along this single dimension than in attempting to 
predict outcomes from any one self -appraisal variable 
independently. Of course, this is best addressed by an 
analysis (e.g., factoring)" (p. 54). 

If Cronbach and Snow (1977) and Corno et al . (1981) 
are correct, stress will be related to self-esteem and 
depression will not be related to self-esteem. 

Dependent Variables 

Right Forebrain Electrical Activity 

Gale (1973), in "The Psychophysiology of Individual 
Differences: Studies of Extroversion and the EEG , " said 



26- 



there have been more than a dozen studies looking at such a 
relationship and they have yielded three classes of out- 
comes. "Extroverts have been shown to be less aroused than 
introverts, more aroused than introverts, or equally aroused" 
(p. 215). He used this as an example of how psychophysio- 
logical studies can be poorly executed because the environ- 
ment or the instructions are not adequately controlled. He 
said "Some sort of task, to which the subject must give 
attention, is essential if the experiment is not to measure 
speed of sleep onset rather than resting EEG" (Electro- 
encephalogram) (p. 224). 

Campbell et al. (1981) in their "Neuroanatomical and 
Physiological Foundations of Extroversion" also presented 
a fairly comprehensive review of extroversion and the EEG. 
They said, "A possible explanation for the contradictory 
reports is that the late "N1-P2" components of the evoked 
potential are influenced by uncontrolled non-sensory factors 
such as attention and motivation" (p. 264). (The N1-P2 are 
standard electrode implacements and are located at the 
higher centers.) Campbell et al. (1981) showed that "ef- 
fects found at higher levels of the brain are probably not 
due to parallel changes in the periphery or the brainstem" 
(p. 263). They found no differences for lower brain level 
EEG among the introverted, ambiverted, or extroverted. How- 
ever, their strict adherence to accepted EEG experimental 
principles was instructive. 



■27- 



Stanley (1982) implied that there would be a functional 
decrease in brain activity in the frontal cortex of ex- 
tremely depressed people (inferred from his suicide autopsy 
studies where he found people who had committed suicide 
had fewer neuroreceptors than those who had not committed 
suicide) . 

Systematic habituation effects were discussed by Rosier 
(1981) in his "Event Related Brain Potentials in a Stimulus- 
Discrimination Learning Paradigm" in the journal Psycho - 
physiology . He manipulated different stages of learning 
and found functionally distinct processes of attentional 
set. He also described four different procedures and ex- 
plained when to use each when using the EEG in experimental 
studies. Another excellent article about the EEG and re- 
search is "The Analysis of Brain Waves" (Brazier, 1962). 
An in-depth discussion of the EEG, electrode implacement, 
and application of computer technology for the analysis 
of the EEG was presented. 

Heart Rate 

Attention to non- threatening external stimuli produced 
heart rate deceleration and hypotension according to Lacey 
et al. (1963). Attention to internal stimuli increased 
heart rate and blood pressure. Breathing rate and skin 
conductance were not affected. Lacey and Lacey (1970) showed 
the same phenomena using heart rate variability, a complex 
procedure differentiating within the heart wave. 



■28- 



Subjects "who are above average in heart rate during 
coping tasks show consistently higher heart rates and sys- 
tolic pressures during other stresses as well, but are 
indistinguishable from less reactive persons when relaxed," 
according to Light (1981, p. 217). In her procedure for 
heart rate during relaxing, she recorded the lowest HR 
elicited. Light also found that those 

with hypertensive parents had significantly 
higher heart rates than subjects with normo- 
tensive parents during both relaxation and the 
avoidance task: however, the two groups were 
most clearly differentiated at the onset of 
the avoidance task when the group with hyper- 
tensive parents averaged 15 beats per minute 
higher than the comparison group. (p. 221) 

Schell and Lusche (1981) reported heart rate differ- 
ences between Type A and B individuals at rest. They also 
found a generally higher SNS "tone" among the Type A sub- 
jects (for blood pressure or other measures taken simul- 
taneously) during their experimental manipulations. 

Van Egeren et al . (1978) investigated whether verbally 
harassed subjects would experience a heart rate increase. 
He noted that heart rate increased with their report of 
anger . 

Appel et al. (1981) investigated between group heart 
rate differences for both high and low blood pressure people 
subjected to anger. There were no heart rate differences 
for anger in the experiment. In fact, heart rate decreased 
although blood pressure increased. 



•29- 



Personality traits (locus of control, state anxiety, 
and trait anxiety) were tested by Levenson and Ditto (1981) 
for predictability of controlling heart rate and nothing 
significant was found. 

Schandry (1981) tested groups of good and poor heart 
rate perceivers and found that the good perceivers had sig- 
nificantly higher scores on a test of state anxiety. He 
then showed that "accurate autonomic awareness is coupled to 
emotional experience and especially anxiety" (p. 475). They 
further said, "it seems that higher self-reported anxiety 
is due to better perception of physiological processes 
rather than to actual level of autonomic arousal" (p. 479). 
Schandry (1981) then seemed to imply a "feedback loop" between 
psychology and physiology. 

Respiration Rate 

Wilier (1980) found a progressive increase in respira- 
tion rate and heart rate as a function of repetition of 
stress in time. These kinds of increases were noted in study 
after study and have been reported earlier in this review 
(Hilgard, 1979, and others). 

In each of the studies in the review of the literature 
presented in previous sections in which respiration rate was 
one of the simultaneously recorded variables, respiration 
rate was not associated with any significant differences for 
any variable (e.g., Stern et al. , 1961). 



-30- 



However, respiration rate affects heart rate and 
generally heart rate increases as breathing rate increases. 
Breathing rate then must be monitored to insure heart rate 
data collection is accurate. An excellent article discussing 
the effect of breathing rate on heart rate variability was 
presented by Mulder and Mulder (1981). Respiration rate 
implications and new technology in psychophysiology were 
discussed. 

Perspiration Rate 

A high degree of arousal, assessed from a number of 
behavioral indexes such as crying and movement, was found 
highly correlated with the galvanic skin response for sixty 
neonates, human two to five day old babies (Weller and Bell, 
1965). 

Schell and Lusche (1981) found significantly higher 
skin conductance for those with Type A personality during 
all of the treatments in their research. They had four con- 
ditions: (1) resting, (2) reaction time, (3) anagram task 
with difficulty varied successively from easy to difficult 
and a loud unpleasant noise sounded at failure, and (4) a 
timed math task with verbal harassment. However, changes in 
skin conductance did not differ between Type A and Type B 
personality subjects for any of the tasks. 

Schandry (1981), as reported previously in this study, 
found that good heart beat perceivers had higher state 



31- 



anxiety. He also found that skin conductance level did 
not change. He speculated this "may be a consequence of 
the rather rapid changes between rest and perception phases; 
possibly this tonic measure was not sufficiently responsive 
to rapid changes, so that the mean values remained un- 
changed" (p. 477) . 

Fenz and Epstein (1967) reported that both novice and 
experienced parachutists have a similar increase in galvanic 
skin response (GSR) as the time for jumping approached. 
However, experienced jumpers reported different timing for 
emotional arousal (just after the jump) while novice jumpers 
reported an earlier arousal (just before the jump) . They 
speculated that "perhaps the experienced jumpers had learned 
to inhibit the subjective response of fear in response to 
the first signs of physiological arousal" (p. 34). 

Hastrup (1979) reviewed literature concerning the rela- 
tionship of vigilance tasks to extroversion/introversion and 
labile subjects, defined as those who have a high frequency 
of spontaneous electrodermal fluctuation and who do not 
quickly habituate. She said many investigators have found 
a relationship between introversion and electrodermal 
lability. She found that introversion was not related to 
electrodermal lability but that it was related to a higher 
initial level of performance. 

Electrodermal responding to aversive stimuli differed 
between depressed and non-depressed persons according to 
many researchers (Gatchel, 1981). Further, many investigators 



•32- 



have found that clinical depression is associated with de- 
creased electrodermal responding (Gatchel, 1981). 

Blood Pressure 

Lacey and Lacey (1970) found that attention to external 
stimuli had a decreasing effect on blood pressure (BP) and 
that attention to internal stimuli had a hypertensive effect. 

Type A personality people generally had higher blood 
pressure than Type B (Schell and Lusche, 1981). They also 
found that a high time pressure task produced greater in- 
creases in blood pressure for Type A personality subjects 
than for Type B subjects. 

Light in 1981 showed that males who have higher than 
average heart rate during coping tasks also have consistently 
higher systolic blood pressure during other stressful tasks. 
However, these subjects were indistinguishable from the 
other subjects when they were relaxing. She found large 
increases for systolic BP but only small increases in dia- 
stolic BP during the stressful tasks. Systolic BP was con- 
sistently more reactive than diastolic BP . 

During the course of her complex experiment Light found 
her various measures did not yield a consistent picture and 
suggested other factors may also be involved. She also 
said "that both higher casual systolic blood pressure and 
high heart rate reactivity to stress are associated with an 
increased incidence of parental hypertension, but high heart 



-33- 



rate reactivity shows a stronger relationship. The incidence 
of parental hypertension is roughly twice as great among 
parents of subjects with mildly elevated casual blood pres- 
sures, but it is almost five times as great among the 
parents of high as compared with low heart rate reactors" 
(1981, p. 222). 

Van Egeren et al. (1978) found that verbally harassed 
subjects had higher systolic BP while solving anagrams. 
After the task, blood pressure reduced back to normal at a 
slower rate when there was uncertainty of consequences com- 
pared with those who were told what to expect next. 

An excellent article giving blood pressure statistics 
in the national population by age, sex, and other relevant 
variables was presented in Hypertension in Adults (1981). 
The summary was particularly instructive as was Appendix 
III (Sources of Variation in Blood Pressure Measurements). 

Between Groups Factors 

Self-Esteem 

Social learning and cognitive variables influence the 
development of the self-concept according to Combs and 
Snygg (1959). "One of the most critical aspects of the 
self-concept is self-esteem" (Mischel, 1976, p. 4). "Self- 
esteem refers to the individual's personal judgement of his 
own worth," according to Coopersmith (1967, p. 8). 



-34- 



Terman and Oden (1959), in their historic longitudinal 
study of the gifted, found that the mortality rate of the 
least successful was twice that of the most successful. 

The above statements, when taken as a whole, imply that 
low self-esteem may lead to earlier death and, by further 
inference, to deleterious physiological variables prior to 
death. 

Additionally the loss of self-esteem can be acute, 
rather than chronic as Terman ' s study of the life span in- 
fers. Burns (1980) said that some people "are likely to 
respond to the perception of failure or inadequacy with a 
precipitous loss in self-esteem that can trigger episodes 
of severe depression and anxiety" (p. 25). 

Bandura (1982) postulated a theory of self -efficacy . 
He said this theory must specify when perceived inefficacy 
will give rise to anxiety and depression. 

Combs and Snygg (1959) talked about the adequate per- 
son and about anxiety and depression. Bandura (1982), al- 
though he would disagree, tested some of Combs and Snygg 's 
principles when Bandura showed "the higher the level of 
self-ef f icacy , the higher the performance accomplishments 
and the lower the emotional arousal" (p. 122). Bandura also 
said that perceived self -efficacy helps to account for the 
level of physiological stress reactions, resignation to 
failure experiences, and achievement strivings, among others 

The "dirty words" studies (McConnell, 1980) showed 
emotional arousal (SNS: blood pressure and heart rate 



•35- 



increases) even when the words were presented at very high 
speeds and were supposedly unintelligible (subjects said 
they did not know what they saw) . Perception of the world 
is done through a f ilter/amplif ier--a feedback loop--the 
self (Bedinger, Bedinger, and Purkey, 1983). McConnell 
(1980) said there "seems to be fairly good evidence that 
something like perceptual defense or vigilance does occur" 
(p. 273). People seem to not always be aware at a cogni- 
tive level of their SNS arousal or the external stimuli that 
produced that arousal. 

Aging 

As people age, there is an increase in individual 
specificity (Garwood and Engel, 1981). This means that as 
people grow older there is an increase in the tendency to 
respond to stimuli with a consistent response hierarchy. 
Physiological measures in the Garwood and Engel (1981) study 
were heart rate, blood pressure, perspiration, breathing, 
and digital blood flow. 

Weg (1975) chronicled age related physiological changes 
in her chapter "Changing physiology of aging, normal and 
pathological" in the text Aging . She mentioned those effects 
generally known such as decreases in brain electrical 
activity, heart rate, breathing rate, and perspiration rate 
and increases in blood pressure. She also said reaction to 
stress decreases with age at SNS stimulation because there 



•36- 



are concomitant changes in hormone levels such as adrena- 
line, noradrenaline, and corticords and thus changes in the 
activity of the organs affected. 

Although reaction to stress decreases, depression may- 
increase according to many researchers (Breslan and Haug , 
1983). Gaitz (1977) proposed that depression is the in- 
evitable consequence of the aging process. However, Bultena 
(1978), in his ten year study, rebutted depression as an 
inevitable consequence when he showed that those with younger 
self-images correlated high with favorable self -evaluations 
and that they were as happy as older people as they had been 
as younger people. 

Wortman and Loftus (1981) further refute inevitable 
depression by reporting that they found the aged either 
extremely happy or extremely unhappy. Their findings also 
seem to concur with the decrease in women's suicide with 
aging ("Vital Statistics of the United States," Mortality , 
Volume II, 1979). This is so because if depression is an 
inevitable consequence of aging one would not be expected 
to find extremely happy older people or decreases in suicide 
rates. However this is a very complex phenomenon since 
many older people are extremely depressed and there has 
been an increase in suicide rate for aging men ( Mortality , 
Volume II, 1979). 

Perhaps one of the most important findings for the aged 
was that active coping strategies increase physiological 
arousal but were associated with lower psychological perception 



■37- 



of anxiety and stress (Miller et al., 1970). This latter 
infers that older people who remain active may suffer less 
anxiety and stress, and with further inference, less depres- 
sion. 

Palmore (1982) in the Duke twenty- five year longitudinal 
study found that several variables were instrumental in 
predicting longevity. For men, health self-rating, work 
satisfaction, and performance intelligence were the strongest 
predictors. Predictors for women were health satisfaction, 
past enjoyment of intercourse, and physical function rating. 
Each of these is somewhat measured by the Tennessee Self- 
Concept Scale. Why then has aging consistently not been 
related to self-esteem as stated by Wylie (1974b) in her 
compendium of self-theory research? 

Breslan and Haug (1983) presented a model that seems 
to describe the path that some elderly take to clinical 
depression. Their model also may predict the answer to the 
above question. Their model involves an interplay between 
developmental changes, special age-related vulnerabilities, 
and the consequences of depression. This model explained 
how many elderly people remain happy while others suffer 
clinical depression. It then also suggests an explanation 
for the consistent lack of correlations between aging and 
self-esteem as reported by Wylie (1974b) because an inter- 
action between aging and depression will mask any relation- 
ship between aging and self-esteem if depression is related 
to self-esteem. 



■38- 



This means then, if there are such relationships, 
healthy older individuals are likely to enjoy high self-esteem 
and less healthy older people will have less self-esteem. 

Pilot Studies 

Many procedures reported in the articles in the above 
review of the literature were investigated in the pilot 
studies. The pilot studies also served as a training 
vehicle for the researcher. 

Thirty subjects took part in the pilot studies. One 
hundred undergraduates were screened using a short- form 
self-concept instrument and ten students from each of the 
high and the low extremes took part in additional research. 
Five graduate students and five older adults also underwent 
the trial procedures. Initial pilot studies might be better 
characterized as single subject designs. The latter pilot 
studies looked at between group differences. 

Many of the subjects were tested two or more times: 
five of the subjects twice the same day, five of the sub- 
jects twice in one week, and five of the subjects every 
month for three months. Reliability of physiological 
responding by psychological treatment was generally high 
for all subjects (r = .60 to .90, p < .01). Test-retest 
reliability of the instruments was very high (r = .85 to 
.95, p < .01) . 



•39- 



The criteria for selection of physiological indices in- 
cluded accuracy, reliability, and the history of the measure- 
ment in the professional journals. Five were selected: 
right forebrain electrical activity measured by the electro- 
encephalograph, heart rate, respiration rate, perspiration 
rate, and systolic blood pressure. The perspiration rate 
was measured independently of the other four and both sys- 
tolic and diastolic blood pressure were measured at the end 
of the debriefing at least ten minutes after the formal 
data collection was completed. These data were taken after 
the subjects thought the experiment was over. The pilot 
studies showed that if the data were collected as a "reward" 
for 'being such a good subject," the subjects were most 
likely to exhibit their "real-world" personality. 

Five psychological variables were selected as treat- 
ments based on the subjects' subjective ability to follow 
instructions, the ability of the physiological measurements 
to discriminate between treatments, and the meaningfulness 
to self-theory suggested by the literature review. 

The five psychological variables selected were mentally 
imaging well being while relaxing, mentally imaging depres- 
sion, machine-produced stress (the subjects in the pilot 
studies reported stress when the blood pressure cuff was 
first inflated), mentally imaging guilt, and instruction- 
produced stress (listening to statements about death). The 
first three (well being, depression, and the stress at the 
first blood pressure cuff inflation) produced the most 



-40- 

reliable results and were used in the primary factorial de- 
signs of the study. The last two, guilt and the "instruction 
stress" that was produced when the subjects listened to 
statements about death, were used to successfully replicate 
the seemingly significant relationships that were found 
between the primary physiological indices and self-esteem 
in the latter pilot studies. 



CHAPTER III 
METHODOLOGY 



The underlying assumption of the present research was 
that one's self-esteem, a psychological state, is directly 
related to one's internal physiology. That is, for instance, 
a low state of self-esteem will be deleterious to one's 
physiology and will disproportionately magnify the negative 
relationship during periods of depression and stress. Fur- 
ther, this adverse relationship, if very meaningful, will 
exacerbate or mask the aging process itself. 

Experimental treatments (psychological events such as 
depression) were produced and measured by dependent variables 
(physiological indices such as blood pressure) . Physio- 
logical indices were then correlated with self-esteem for 
each psychological event to see if a relationship was pre- 
sent . 

Subjects were next divided into groups by self-esteem 
so that treatment by self-esteem interactions could be 
assessed, and later into younger and older age groups in 
order to assess the practical as well as the theoretical 
implications of the research. Practical implications were 
gained in two ways. One way was to take the younger and 
older age groups and simply look at the "real-world" data 



-41- 



•42- 



like that which had been produced in the pilot studies when 
the subjects' blood pressure was given as a "reward" for 
"being such a good subject." The other was to assume that 
if self-esteem is a meaningful construct, self-esteem effects 
will produce predictable confounding (e.g., masking) when 
comparing the younger and older age groups. For instance, 
the older age group should have higher blood pressure. But 
if there are more younger people and their blood pressure (BP) 
increases at a greater rate (this predictable phenomenon is 
explained in the next four paragraphs below) , there should 
be no BP differences between the younger and older age groups 
even though the older age group normally has higher casual 
blood pressure. 

The factorial design of the present study precluded an 
investigation of the interaction between younger and older 
age groups by self-esteem because few older people were ex- 
pected in the two extreme self-esteem (se) groups (high and 
low se). Design cells with extremely small numbers are 
not at all reliable. Some masking and exacerbating effects 
were predictable, though, because aging characteristics 
are highly reliable and directional. The five physiological 
variables selected in the present design either increase or 
decrease with age. That is, for instance, blood pressure 
increases and brain electrical activity decreases with age. 

This would mean then that an older age group, already 
suffering from adverse increases in a physiological vari- 
able such as blood pressure, will have an average BP that 



•43- 



will no longer exceed a younger age group mean when the 
younger group suffers disproportionately higher blood pres- 
sure during experimental treatments (assuming more younger 
people than older people suffer disproportionately higher 
blood pressure) . This is so in the present experiment 
because, with very small numbers of older people expected 
in the high and low se groups, any effect for se will be 
borne mostly within the younger group. Stated in another 
way, if there is a disproportionate significant increase in 
blood pressure for those with low self-esteem and the low 
self-esteem group is composed mainly of younger people, 
the overall younger group mean will increase enough to 
mask an a priori difference between age groups if the effect 
for se is very meaningful. 

Therefore, there will be no statistical difference 
between the younger and older age group means as measured 
by a directionally increasing variable, blood pressure, 
during psychological events that also increase that measure, 
such as depression and stress. 

However, should there be a priori differences between 
the two groups of younger and older people in the other 
(negative) direction (e.g., brain electrical activity 
decreases with age), the differences would be magnified. 
This is so because the younger group, if low self-esteem 
people's average output is disproportionately increased 
by depression or stress, will have their already higher 
output exacerbated. 



•44- 



Four dependent measures that decrease with age were 

selected from the many variables that were pilot tested: 

right forebrain electrical activity, heart rate, respira- 
tion rate, and perspiration rate. 



Hypotheses and Corollaries 

The above assumptions and inferences generated the 

following hypotheses and their corollaries: 

H, n : Right forebrain electrical activity will 
be negatively related to self-esteem. 

C-, i : People with low self-esteem will have rela- 
tively higher right forebrain electrical 
activity than those with high self-esteem 
during states of depression and stress. 

C-, ^ : Older people will have lower right fore- 
brain electrical activity than younger 
people during states of depression and 
stress (exacerbated EEG decrease) . 

H~ (-, : Heart rate will be negatively related to 
self-esteem. 

C~ , : People with low self-esteem will have a 

higher relative heart rate than those with 
high self-esteem during states of depres- 
sion and stress. 

C~ ~ : Older people will have a lower heart rate 

than younger people during states of depres- 
sion and stress (exacerbated HR decrease) . 

H~ n : Respiration rate will be negatively related 
to self-esteem. 

C„ ■, : People with low self-esteem will have a 
relatively higher respiration rate than 
those with high self-esteem during states 
of depression and stress. 



■45- 



Co 2 : Older people will have a lower respiration 
rate than younger people during states of 
depression and stress (exacerbated RR de- 
crease) . 

H, Q : Perspiration rate will be negatively related 
to self-esteem. 

C, , : People with low self-esteem will have a 
relatively higher perspiration rate than 
those with high self-esteem during states 
of depression and stress. 

C, ? : Older people will have a lower perspiration 
rate than younger people during states of 
depression and stress (exacerbated GSR de- 
crease) . 

H,- « : Blood pressure will be negatively related 
to self-esteem. 

Cr , : People with low self-esteem will have 

higher relative blood pressure than those 
with high self-esteem during states of de- 
pression and stress. 

Cr ? : There will be no blood pressure differences 
between younger people and older people 
during states of depression and stress 
(masking effect) . 



Subjects 

Fifty-five volunteer Caucasian adults were recruited 
from the Unitersity of Florida and Lake City Community 
College. Of the 55, two subjects were older adults. Addi- 
tionally, fifteen volunteer Caucasian older adults were 
recruited through the Gainesville, Florida, Older Americans 
Council (total N = 70). 

The age range of the younger age group (N = 53) 
was 15-59 years and the older age group (N = 17) was 



■46- 



60-79 years. Older people were then one-fourth of the total 
sample. 

Thirty-five of the subjects were in the upper- lower 
socio-economic status (SES) group. Twenty-one were lower- 
middle SES and fourteen were upper-middle. Ten subjects 
had previously undergone psychological therapy. The sub- 
jects' SES and/or previous therapy history were thought to 
be possible confounding variables but they were not. This 
is shown in Appendix I where statistical analyses of demo- 
graphic variables are presented. It shows that the subject 
population was not statistically different from the 
general population except that their average age was 
higher since the research design called for comparison 
of a younger and an older age group. 

Two subjects were eliminated from the results, one 
Black person because enough Black volunteers were not forth- 
coming and one Caucasian because she refused to complete 
the Tennessee Self-Concept Scale after data collection on 
the physiograph. 

Subjects were not asked to participate if they were 
more than plus or minus one standard deviation from the 
weight norm for their height or if they had taken any drugs 
within twenty-four hours of data collection. Only healthy, 
normotensive subjects were recruited. 



-47 



Apparatus 

Three separate apparati were used of which two, a 
NARCO Bio-Systems Phsyiograph and a Lafayette Student 
Galvanometer, were used simultaneously. The physiograph 
consisted of the following systems: 

PMP-4B Physiograph with five second and event 
marker and four Channel Amplifiers, Type 
7070 

Programmed Electro-Sphygmomanometer , PE-300 
with occluding cuff Transducer Coupler, Type 
7173 

Hi-Gain Coupler, Type 7171 (2 each) 

Impedance Pneumograph Coupler, Type 7212 

A Standard K-085 cuff blood pressure measurement ap- 
paratus was used independently in time from the other two 
(used at exit interview instead of during formal data 
collection) . 

Instrumentation 

The Tennessee Self-Concept Scale was used to measure 
self-esteem and the Hollingshead Two-factor Socio-economic 
Status Scale was utilized to measure socio-economic status 
(SES) . 



■48- 



Procedure 

The subjects were requested to read and sign two in- 
formed consent forms (Appendix II) and retain one since they 
were considered to be at risk by the University of Florida 
Human Subjects Committee. (Any research involving human 
subjects, where their physiology is being monitored by elec- 
trodes attached to equipment with potentially harmful amounts 
of electrical current anywhere in the system, constitutes 
an "at risk" condition. See Appendix III). Subjects next 
completed a medical history and SES questionnaire (Appen- 
dix IV) . 

Laboratory Setting 

The laboratory consisted of a desk area facing a wall 
where subjects completed the above mentioned forms. The 
other side of the room was a blank wall, virtually stimulus 
free, and the physiograph apparatus stood in the center of 
the room behind a reclining chair. 

Subjects were seated in the reclining chair. They faced 
the blank wall and the physiograph equipment was behind them. 
The physiograph operator (the researcher) stood behind the 
physiograph but in a position to closely observe the subjects 
and to record observations on the chart paper as it moved 
across the machine. 

Electrodes were implaced by a same sex graduate student. 
Three were placed in the sternum area (care was taken to 



-49- 



insure all electrode implacement was identical for all sub- 
jects) for recording of heart rate (ECG) and respiration 
rate. Two finger electrodes were placed on the first two 
fingers of the right hand for recording the perspiration 
response. Subjects were asked to report if there was any 
pulse throbbing in the finger tips and, if so, electrodes 
were loosened. Subjects' comfort was stressed throughout 
data collection. Two electrodes were placed under a head- 
band one-half inch above the right eye to measure right 
forebrain electrical activity (EEG) . The blood pressure 
(BP) cuff was affixed to the left upper arm and the physio- 
graph turned on. 

Standard time constants for human subjects for the ECG 
(3.2) and EEG (0.3) were used. Respiration rate settings 
varied by subject to insure that both normal and stressful 
breathing rates were measured. Directions from the Narco 
Physiograph Manual (1980) were scrupulously followed for 
each measure. 

Instructions 

After ascertaining that heart rate, respiration rate, 
perspiration rate (GSR) , and electroencephalograph were 
being monitored and recorded correctly while subjects were 
relaxing (see Appendix V), the researcher stated in a carefully 
practiced delivery, "I'm going to turn the blood pressure 
cuff on now. Let me know if you experience any discomfort." 



-50- 



This statement was found to remove a "startle response." 
The initial BP cuff inflation often produced such an effect 
in the pilot studies. The not too subtle implication that 
discomfort was expected was then timed with the BP cuff 
inflation, beginning at cuff inflation and ending at fifty 
millimeters of Mercury (mm Hg) pressure. The cuff continued 
to 150 mm Hg, well below the 180 to 200 mm Hg typically de- 
livered by a nurse or medical doctor and then deflated at 
the same rate. The machine repeatedly inflated the BP cuff 
twice per minute and BP was recorded twice each inflation 
(four times per minute). 

Subjects were again requested to relax so that the 
physiological indices would return to the baseline for the 
recording of the well being state (see Appendix V). Data 
were then collected while subjects imaged depression (see 
Appendix V) for at least two minutes. Again, a practiced 
delivery was accomplished during the mentally imaging 
states. Appendix V operationalizes imaging states. 

Next, as an intervening task, twelve statements from 
the Tennessee Self-Concept Scale and eight additional self- 
concept statements were read while the subjects relaxed and 
listened. Fifty-three of the subjects selected randomly 
were then asked to mentally image guilt (Appendix V) . 
Twenty-two subjects were also requested to listen to state- 
ments about death (mentally imaging a stressful event, 
Appendix V). All subjects imaged well being again as their 
last task on the physiograph. 



51- 



After completing data collection, the researcher re- 
moved the electrodes and BP cuff and requested the subject 
take the Tennessee Self -Concept Scale. Identification of 
high and low self-esteem groups was not done until all raw 
data had been evaluated and recorded. 

After finishing the paper and pencil task the subjects 
were given an exit interview and their blood pressure was 
taken by the traditional method. This was offered as a 
"reward" for "being such a good subject." This is also, 
of course, additional data that may be more "real-world" 
than the "laboratory" data that had just been collected. 
The "laboratory" data were probably slightly confounded by 
"anticipation" while the "exit" data were less likely to be 
so influenced. "A priori" blood pressure was inferred from 
this "real-world" resting blood pressure (the older age 
group was expected to have higher a priori blood pressure) . 

General Design Used in the Present Experiment 

The above procedure resulted in five 3x3 factorial 
designs each with an N of 70 (self-esteem group by treatment 
for each dependent measure). The additional 53 subjects in 
the "guilt" condition and the 22 in the "image stress" con- 
dition created an unequal-cell 3x2 (self-esteem group by 
treatment) for each of the five measures. This gave the 
added ability to statistically compare two mentally un- 
pleasant images (depression and guilt) and two stressful 



•52- 



events (one machine-produced at the first BP cuff inflation 
and the other instruction-produced while listening to state- 
ments about death). This gave test-retest reliability. 

Figure 1 is a conceptual integration of these designs. 
Figure 1 shows each psychological state was reflected 
separately by the different physiological variables. These 
were further divided into self-esteem groups. For instance, 
70 subjects imaged well being as measured by the electro- 
encephalogram and 19 of these were in the low self-esteem 
group. 

Figure 2 is the design used for looking at age group 
differences. Younger and older age group means were 
analyzed in a 2 x 3 (age group by psychological treatment) 
factorial for each dependent measure (physiological vari- 
able). They were further analyzed for replication of sig- 
nificant findings in a 2 x 2 factorial for the physiological 
variables found significant for age group by psychological 
treatment in the prior 2x3 factorial. 



•53- 




C-4 CTv 


DO! r^ on 


>> • 




r-tn CNI f-H 




XI 


-o 




to ,1 


•on 2 II 


> • 


•H 


■H O 


C ■•"! • 


X z 


X S J 2 


•H U ■ 

ra o • 
u< ■ 






W32J 


S3-J'I3S 


u a ■ 

o o ■ 

u. -^ ■ 

u ■ 

£. U O 

00 "J UJ 
•HHUJ 
OS Id — 



6 

•H 

a 



e 
aj 
w 

a 



a 

60 
02 

T3 



60 
■H 



saiaviHVA xMHc:;3a3a 



-54- 



u 

T3 



S3U3U193B3S 

SufasaH 




S3'iqviHVA j.NHai^'id'ia 



X5 

C 

CO 
01 

M 
g 

o 
c 

QJ 

•p 

qj 



0) 

o 
C 

OJ 

QJ 
4-1 
IW 
•H 



4-1 

t) • 

QJ W 

cn Cu 

o 

a n 

GO M 

•r-l 

W QJ 

QJ M 

Q cd 



QJ 

M 
•H 
ft, 



CHAPTER IV 
RESULTS 



Subjects selected for the experiment did not differ 
significantly from the general population except in expected 
design-determined directions. This was so because there 
were more people over age 60 in the sample than in the 
general population so that a comparison between a younger 
group and an aged group would be possible (e.g., this older 
group had average higher systolic blood pressure at exit) 
(see Appendix VI). 

Demographic Results 

There were no significant relationships between age 
and self-esteem as measured by the Tennessee Self-Concept 
Scale (TSCS) (r = .08, p < .05). When the subjects were 
divided into three groups by self-esteem (high self-esteem 
was defined as those in the sample who scored in the upper 
27% on the TSCS; the middle self-esteem group was the center 
46%; and the low self-esteem group was the bottom 27%), the 
middle group age mean was significantly higher (F(2,69) = 
3.19, p < .05) than the other two groups. This difference 
was of course predicted since a higher percentage of subjects 
over age 60 in the middle group was expected. Of the 17 
subjects in the over 60 age group, three fell in the highest 



-55- 



•56- 



27%, four in the lowest 27%, and ten in the middle group, 
which elevated the middle group age mean. 

The high self-esteem group age mean was 35.5 years 
(N = 19, a = 14.9), the middle self-esteem group age mean 
was 45.6 years (N = 32, a = 18.8), and the low self-esteem 
group age mean was 34.0 years (N = 19, a = 18.1). 

Post hoc analysis using the Duncan procedure (DF = 67, 
MS = 326.64, p < .05) confirmed that the age of the middle 
se group was significantly different from both the high and 
low se groups while the high and low groups were not signifi- 
cantly different from each other for age. Thus any differ- 
ences found for measured states of high and low self-esteem 
are more meaningful for self-esteem comparisons with the 
middle group removed because the middle group physiology was 
influenced by the larger percentage of aged people within 
that group. However, graphic results which include the 
middle group do demonstrate age comparisons because the 
middle group has a disproportionate share of older people. 

Given that the demography of the subject sample was not 
identical to the overall national population, there were 
still no significant relationships between self-esteem and 
the demographic variables measured: age, sex, socio-economic 
status, and whether or not the subject had had psychological 
therapy (see Appendix I) . 

Further, there were no significant demographic differ- 
ences within the groups of measured states of high and low 
self-esteem (see Appendix I) . 



-57- 



Results of analyses of psychophysiological measures 

(dependent variables) follow and are again organized into 

sections in the following order: 

Electroencephalogram 
Heart rate 
Breathing rate 
Galvanic skin response 
Blood pressure 

Results within each of these sections are reported in 
the following order: first, the correlations between the 
dependent measures (e.g., blood pressure) and self-esteem; 
second, any differences between measured states of self- 
esteem; and third, any differences between younger and older 
age groups. 

Right Forebrain Electrical Activity 

Electroencephalogram (results of analyses of psychophysio- 
logical relationship between self-esteem and right forebrain 
electrical activity as measured by physiograph (electro- 
encephalogram) during experimental treatments) 

Analysis of variance of right forebrain activity (EEG) 
showed that mentally imaging well being, imaging depression, 
and the mental event taking place at the first BP cuff in- 
flation (stressful event) by the physiograph were signifi- 
cantly different from each other (F(2,134) = 41.17, p < .001) 
Further, the interaction between self-esteem and the treat- 
ments was significant (F(4,134) = 4.80, p = .001) (see 
Table 1) . 



fn 



w 


r-~ <r vo <r 


S 


r- co <r md 


<J 




p 


nciooo 


C 


o-\ 



°S 
en o 
W O 
w w 

O Pd 

o 



o o 
o o 



oo r*. 
oo cm 



O H 
O CN 



rH cni r^ CN1 









C 












o 












•H 












4-1 












cfl 












bO £ .-1 












C O M-l 












•H -r-l C 


CQ 




CX 






0) CO -H 


ex 




3 






XI to 


3 




o 






1 01 U-l 


O 




u 






rH M M-l 


M 




bO 






.-1 (H, 3 
QJ 0) O 


bO-U 




E 






5 T3 


E oi 




OJ 






PL, 


01 B 




01 




4-1 


bO bOpq 


01 J-) 




CO 4J 




C 


c c 


■U rd 




4-) CO 




') 


•H -H 4-) 


CO 0) 




a a) 




B 


bO bO co 


cu n 




CD 1 


M 


u 


c3 cfl U 


1 -U 


M 


•r-lM-l 





en 


G E -H iw 





XI cH 


^ 


i> 


•H T-l 4-( 


r-l X 


u 


3 o> 


Vh 


u 




0) 


u 


en en 


Jj 


H 




en 


LU 



bfl 



o o 



v£) CO 
CO v£> 



co <r 
o o 



co r-« 
o o 



I— I I— I Csl 







LM 








d 


C 


bO C 


ej 


o 


C 







•H 


•H 


•H 


Pm 


-U 


ai 


CO 


pq 


CT3 


-Q 


CO 




r— ( 




0) 


4-1 U-l 


i—i 


1-1 


CO 


c 


rH 


a, l-i 


•r-l 


OJ 


0) 


■t-l 




3 O fc 





59- 



H, n : Right forebrain electrical activity will be 
negatively related to self-esteem. 

Right forebrain activity measured by an electroencephalo- 
graph (EEG) was positively correlated with self-esteem (r = 
.31, p < .01) while subjects were imaging well being (relax- 
ing). However, imaging depression showed no relationship 
(r = .09, p > .05) with the TSCS. The mental event at the 
first BP cuff inflation (stressful event) was negatively 
correlated with the score on the TSCS (r = -.26, p < .05). 

This result was replicated when twenty-two subjects 
who listened to statements about death (also a stressful 
event) had a negative correlation (r = -.41, p < .05) be- 
tween their EEG and self-esteem. 

The finding of no relationship between the EEG and 
depression was not replicated. Fifty-three subjects who 
mentally imaged guilt (a mentally unpleasant event similar 
to depression) were found to have their EEG and self-esteem 
related (r = -.41, p < .05). 



!-, -, : People with low self-esteem will have relatively 
higher right forebrain electrical activity than 



C- 

those with high self-esteem during states of 
depression and stress. 



During the well being state the self-esteem group means 
were not significantly different from each other (Kruskal- 
Wallis chi-square = 1.72, DF 2, p > .05). Therefore, while 
the subjects were relaxing, their right forebrain activity 
tended to be positively related to their score on the TSCS, 
but not enough to show statistical differences between the 
high and low self-esteem groups. 



•60- 



The Kruskal-Wallis chi-square non-parametric statistic 
was significant (chi-square = 9.67, DF = 2, p < .01) for the 
self-esteem groups when the subjects were mentally imaging 
depression. The middle self-esteem group with the dispro- 
portionate number of older people had a significantly lower 
EEG (1.28, a = 0.4) while imaging depression (Wilcoxon t-test 
approximation 0.43, p < .05). High (1.63, a = 0.4) and low 
(1.94, o=0.7) self-esteem groups showed no differences. 

The Kruskal-Wallis test indicated a reliable difference 
for the self-esteem groups (chi-square = 7.67, DF = 2, 
p < .05) at the first BP cuff inflation stressful event 
(see Figure 3) . 



1 1 ? : Older people will have lower right forebrain 
electrical activity than younger people durii 



\ty than younger people during 
states of depression and stress (exacerbated 
EEG decrease) . 



Older subjects did not differ from younger subjects 
when mentally imaging well being or when mentally imaging 
depression. However, the event at the first BP cuff infla- 
tion showed significant differences between older and 
younger people (Wilcoxon t-test approximation 0.68, 
p < .05). 

The older subjects' mean EEG at the first BP cuff in- 
flation (stressful event) was 1.58 (o = 0.6, N = 17), and 
was significantly lower than the younger group mean EEG of 
1.96 (a = 0.6, N = 53). Aging differences were exacerbated 
during stress. 

An attempt to replicate this finding with a small sample 
of 22 subjects during a different stressful psychological 



■61- 



EEG 
activity 



2.5 
2.0 



1.5 



V 



^* D 

,. W 



i.oi- 



0.5- 



low middle high 
SELF-ESTEEM 



Figure 3. Right forebrain electrical activity during 

experimental treatments and its relationship 
to self-esteem. (Note: Middle group had 
disproportionate number of older subjects and 
age was negatively associated with EEG activity 
resulting in depressed middle group means.) 

First BP cuff inflation (stressful event) (S) 

Imaging depression (D) 

Imaging well being (W) 



■62- 



event (listening to statements about death) was not success- 
ful but was suggestive that a larger N would be significant 
(Wilcoxon t-test approximation 0.13, p = 0.07). 
When listening to statements about death (reported as stress- 
ful by 91% of the subjects), the older age group mean EEG 
was 1.33 (o = 0.3, N = 9), and the younger group mean was 
1.85 (a = 0.3, N = 13). Thus older subjects most probably 
displayed lower EEG output again and this subtractive effect 
exacerbated differences since the treatments disproportion- 
ately added electrical output to the younger group mean. 

Heart Rate 

Electrocardiogram (results of analyses of psychophysiological 
relationship between self-esteem and heart rate measured by 
physiograph during experimental treatments) 

Heart rate in beats per minute varied significantly 

between treatments (F(2,134) = 27,05, p < .001) (see Table 

2). The Duncan statistic (DF = 134, MS = 18.03, p < .05) 

showed the three treatments were different from each other. 

H~ n : Heart rate will be negatively related to self- 
esteem. 

There were no significant relationships between self- 
esteem and the experimental treatments when measured by 
heart rate (HR) in beats per minute. The Tennessee Self- 
Concept Scale (TSCS) by HR while imaging well being was 
r = .01 (p > .05), the TSCS by HR for depression was r = 
-.07 (p > .05), and the TSCS by HR at the first BP cuff 
inflation was r = -.02 (p > .05). 



•63- 



fe 



w 

2 Pi 

<<; 

w P 

CO 



°S 
co o 

w o 
w w 

Pi w 
O Pi 

p 



O r^ 



o 



o 



m o 
oo 



O CM VO CO 

cm i— i <r r-» 

<£> r— I r— i oo 



vo co 
co o 



r-l CM 1^ CM 









c 












o 












•H 












4-) 












CO 












M C rH 


X 










c o m 






1X1 






•H -H C 


ra 




ex 






a) en -H 


ex 




3 






X w 


3 




o 






(U 4-4 


o 




v-l 






H MtH 


5-4 




M 






H £X3 

ai a) a 


bfl 




B 






3 t3 


B 4-1 




(U 






Ph 


QJ C 




CD 




4-> 


6C Mpq 


<D 0) 




U3 4-1 




C 


C C 


4-J Ed 




4-> W 




(1) 


•H -H 4-1 


W 4-) 




O OJ 




e 


bDMin 


CD CO 




<U 1 


i-t 


4-1 


co co w 


1 0) 


S-i 


• t— )4-l 


o 


fl 


6 E -h 


4-1 S-i 


o 


XI r-l 


Vj 


'J 


•H -H 14-1 


i— 1 4J 


M 


2 a) 


M 


V-i 




OJ 


J-l 


L0 CO w H 




to 


UJ 






CM CO 



On <r 
^o r-- 



o m 



<r cm 
a> cm 






z 




<c 


4-1 


f T 1 


4-1 


§ 


3 c 




MC OO 


H 


CO -H 


S3 


•H -r-l P-i 4-1 


i 


0) co pq to 


XI M i—l 


H 


<y 4-1 4-4 


< 


<— I U co C 


W 


H CV4-H 


Pi 


ai cu -h 


H 


SQfa 



-64- 



C 9 -, : People with low self-esteem will have a higher 
relative heart rate than those with high self- 
esteem during states of depression and stress. 

During the three treatments there were no differences 

between the self-esteem groups (F < 1, p > .05). 

C~ ? ■. Older people will have a lower heart rate than 
younger people during states of depression and 
stress (exacerbated HR decrease). 

There were no significant differences between the 
younger and older age groups during the experimental treat- 
ments with heart rate as the dependent variable either. Thus 
well being (F < 1, p > .05), depression (F = 2.19, p > .05), 
and the event at the first BP cuff inflation (F < 1, p > .05) 
provided no heart rate differential between younger and 
older age groups. Exacerbated HR decrease was not found. 
However, any differences found for age would not be rele- 
vant for this research since HR was consistently found not 
related to self-esteem. 

Respiration Rate 

Respiration rate (results of analyses of psychophysiological 
relationship between self-esteem and respiration rate 
measured by physiograph during experimental treatments 

Respiration rate in breaths per minute varied signifi- 
cantly (F(2,134) = 4.52, p < .01) between treatments (see 
Table 3) . 

Ho „: Respiration rate will be negatively related to 
self-esteem. 

Correlations between self-esteem and the two mental 

imaging conditions (well being and depression) were not 



■65- 



p-, 
co 
w 
Pi 

O 

X 
Q fa 



Pi O 
O O 
fa H 
pej co 

en w >> 

fa re 

W fa 

pa o jh 

<2FQ 
M Q 

Pi w 
<J Pi 

> D 

CO 
fa < 



fa 



W 
J3 Pi 

CO 



fa 

o 

CO o 

W Q 

fa W 

Pi fa 

o 3 

fa fa 
Q 



o co 

O r-l 



r-l <r 

LO O 



m csi 

i— i 

00 



CO in i— I i— I 

~a- <r vo m 
ct\ i— i m <r 

r^ i— i 



cm r^ cm 






CO 









C 












o 












■H 












4-) 












CO 












MCH 


x 










C O IH 












•H -H C 


CO 




Ch 






0) CO -H 


a 




3 






XI CO 


^ 




O 






CU U-l 


o 




U 






i-i m m 


u 




M 






r-i a, d 
cu a) a 


bO 




6 






£ T3 


E *-> 




OJ 






fa 


a) c 




0) 




4J 


M bOcq 


a) a) 




1/1 4-1 




c 


C c 


4J H 




4-1 CO 




(1) 


•H -H U 


CO 4J 




O QJ 




H 


bO bO co 


QJ cfl 




<U 1 


H 


4J 


cfl co }-i 


1 CU 


u 


•i — i M— 1 


o 


cO 


E E -h 


U-l M 


o 


X r-l 


u 


tu 


•h <h m 


i— 1 4-1 


U 


3 a) 


u 


M 




CU 


u 


CO CO 


w 


H 




00 


w 



x 






v£> CM 
CO CO 



r^ o 
in r-~ 



CO CO 

m vo 



<r co 



<r co 



co o 
co m 



'4-1 
M-l 

3 c 
60 d a o 

CO -H 

•rl >H Pj 4J 

QJ CO PQ CO 
XI CO i—l 
0) 4-1 M-4 
r-l U CO £ 
r-l (X U -rl 

a) cu -H 

!2 Q fa 



•66- 



significant (respectively .08 and .06, p > .05), but the 
relationship between the TSCS and respiration rate at the 
first BP cuff inflation was significant (r = .26, p < .02). 
However the TSCS by another psychologically similar (stress- 
ful) event (hearing statements about death) was not related 
(r = -.11, p > .05). 

Co -, : People with low self-esteem will have a rela- 
tively higher respiration rate than those with 
high self-esteem during states of depression 
and stress. 

There were no differences between self-esteem groups 

(F(2,67) = 2.04, p > .05) during the three experimental 

treatments. Interestingly, all nine means were directional 

from low se to high se. 

C„ ? : Older people will have a lower respiration rate 
(RR) than younger people during states of de- 
pression and stress (exacerbated RR decrease). 

There were no significant differences between the 
younger and older age groups while mentally imaging well 
being and depression with respiration rate as the dependent 
variable (F < 1, p > .05 for both). 

However the older age group took significantly (F(l,68) 
= 5.07, p < .05) more breaths per minute at the physiograph ' s 
first BP cuff inflation (stressful event). 

The younger age group's mean respiration rate was 14.5 
(a = 5.0, N = 53) breaths per minute, and the older was 17.4 
(a = 3.9, N = 17) . 

An attempt to replicate this finding during a different 
stressful event (listening to statements about death) was 
again not successful (F < 1, p > .05). 



•67- 



Respiration rates between age groups were not exacer- 
bated during states of depression and stress as predicted. 
However, the expected a priori difference that older age 
groups have a lower respiration rate was also not found. 



Perspiration Rate 

Perspiration rate (results of analyses of psychophysiological 
relationship between self-esteem and perspiration measured 
by galvanic skin response meter independently of the physio- 
graph, but during experimental treatments) 

The experimental treatments were significantly different 

from each other when measured by the GSR (F(2,134) = 53.28, 

p < .001) (see Table A) . 

H, n : Perspiration rate will be negatively related to 
self-esteem. 

There were no significant correlations between self- 
esteem and the three experimental treatments as measured by 
the galvanic skin response (GSR) . The TSCS and well being 
were thus not related: r = . 06 (p > . 05) . The TSCS and 
depression r = .09 (p > .05) and the TSCS and the mental 
event at the first BP cuff inflation r = .18 (p > .05). 

C, -, : People with low self-esteem will have a rela- 
tively higher perspiration rate than those with 
high self-esteem during states of depression 
and stress. 

There were no differences between the self-esteem groups 
(F(2,67) = 1.58, p > .05) during the experimental treat- 
ments . 

C, 9 : Older people will have a lower perspiration rate 
than younger people during states of depression 
and stress (exacerbated GSR decrease). 



■68- 



>■ 




H 




M 




fa 


i— i 


M 


O rH 


PQ 


O CN 


< 




PQ 


O O 


O 




Pi 




Pm 





fa 



w 
2 Pi 
<:< 
w o 
so- 

en 



fa 

o 
S 

CO o 
fa Q 

w w 

Pi w 

O Pi 

fa fa 
Q 



m oo 
oo m 



oo 
cn 



cr><r r>N oo 


en o> 


cn ^o vo r^ 


<r co 


VO CM «^> a\ 


CO CTv 


o <r o oo 


i—l i— 1 


Nvo<rvo 


CO CN 


vO i— 1 




<r i-h 











c 












o 












■H 












+J 












cd 












bO c i-h 


X 










com 






en 






•H -H d 


CO 




a. 






0) CO -H 


Cu 




3 






,Q CO 


3 




c 






CD 4-1 


O 




M 






rH 5-1 4-1 


H 




to 






H P.? 
QJ 0) O 


W) 




B 






g T3 


e j- 1 




0) 






Ph 


oj c 




cu 




u 


60 bOPO 


a) a) 




co u 




C 


C C 


4J la 




4-J CO 




a) 


•H -H 4-1 


CO 4-1 




O CO 






W) 50 CO 


<U CO 




01 1 


5-1 


u 


Cd CO (-1 


I 0) 


h 


■r-)l4-| 





a 


e e-H 


CM )-( 





XI rH 


u 


cu 


•M -H 14-1 


rH -U 


M 


d <D 


u 


Vj 




0J 


I-H 


co en u H 




co 


UJ 



CO 


ro r^ 


CTi 




r~» m 


oo 


J= 






M 


-d- i-h 


l*» 



o o 



o-> o 
m m 



oo in 
md o 



CO 




^ 




<; 


U-l 


fa 


CM 


^ 


3 C 




GO C O O 


H 


CO -H 


2: 


•H -H P-l 4J 


£ 


0) CO PQ cfl 


XI CO i—l 


H 


CU 4-J CM 


< 


i—l M CO C 


W 


rH Cu U -H 


Pi 


0) 0) -rH 


H 


3: Q fa 



-69- 



While mentally imaging well being there were no dif- 
ferences between the younger and older age groups (F < 1 , 
p > .05). However, while imaging depression and at the first 
BP cuff inflation, the older subjects had significantly lower 
GSR. Both findings were significant during replication with 
psychologically similar events (guilt and stress) but any 
GSR differences for age are not relevant to the present 
dissertation because the GSR was consistently found not 
related to self-esteem. Findings of an exacerbated GSR 
decrease were not then possible with the present design. 

Blood Pressure 

Systolic blood pressure (results of analyses of psychophysio- 
logical relationship between self-esteem and systolic blood 
pressure measured by physiograph during experimental treat- 
ments) 

Systolic blood pressure (SBP) was recorded on the 
physiograph chart while the subjects were mentally imaging 
well being, mentally imaging depression, and when the blood 
pressure cuff was first machine inflated (first BP cuff 
inflation) . 

These three treatments (imaging well being, imaging 
depression, and the first BP cuff inflation) were signifi- 
cantly (F(2,134) = 74.79, p < .01) different from each other 
(see Table 5). Further, the se groups by treatment inter- 
action was significant (F(4,134) = 3.76, p < .01). 



-70- 



fa 



w 
z fa 

< < 

CO 



fa 
o 

co o 

fa Q 

W fa 

Pi fa 

O fa 

W fa 
Q 



O i— I 

o o 



CN O 



O 00 00 r— I 
O 00 00 00 



<r en i— i cn 



<r o 
i— i cn 









a 












o 












•H 












4J 












CT3 












bfl C rH 


X 










C O <W 






w 






•H -iH C 


CO 




a 






cu co -h 


CX 




3 






,n co 


3 




o 






QJ 14-1 


O 




u 






■-I U M-l 


M 




W) 






h a3 
cu cu a 


txO 




e 






15 13 


R 4-> 




aj 






fa 


a) c 




QJ 




4-) 


ojo torn 


QJ QJ 




co u 




c 


c c 


w h 




4-) co 




a 


■t-l -H 4-1 


CO 4-1 




a a) 




R 


W) W) co 


QJ cfl 




0) 1 


M 


4_) 


n3 cd U 


1 QJ 


M 


•i— )U-I 


O 


cn 


R R -H '4-i M 


o 


jQrH 


1^ 


C.J 


• r-l -H M-l 


r-H 4-) 


u 


d <d 


^ 


5-i 




CD 


u 


co co W H 




CO 


UJ 



bC 
•H 



QJ 


cn r-- 


LO 


i— 1 


O CN 


o 


T) 






T3 


cn cn 


CM 



oo cn 
o o 



cn ^o 
cn vo 



00 oo 
o o 



CO 


vo in 


cn 




O^ CTv 


cn 


s 






o 


CN i— 1 


r— 1 







3 


c 


00 C 


o 





C 







•H 


•H 


•H 


fa 


4J 


QJ 


CO PQ 


CI) 


-n 


CO 




r-i 




QJ 


4-1 <4-l 


r-< 


Vj 


CO 


c 


r-l 


CL U 


•r-t 


QJ 


CI) 


•H 




3Qt. 





-71. 



H s n : Blood pressure will be negatively related to 
self-esteem. 

The relationship between well being and self-esteem 
(measured by the TSCS) was not significant (r = .07, p > .05) 
However, the correlations between the TSCS and mentally 
imaging depression and the mental event at the first BP cuff 
inflation were statistically significant (r = -.22, p = .05 
and r = -.27, p = .02, respectively). 

Replication of the finding for depression was accom- 
plished by having fifty-three subjects mentally image guilt 
(also a mentally unpleasant event). The SBP was related to 
the TSCS (r = -.38, p < .05) . 

Further replication of the finding for stress (first 

BP cuff inflation) was suggestive that a larger number of 

subjects would be related. Twenty-two subjects listened to 

statements about death (also a stressful event) and their 

SBP was related to their self-esteem (r = -.39, p = .06). 

C s -, : People with low self-esteem will have higher 
relative blood pressure than those with high 
self-esteem during states of depression and 
stress . 

Systolic blood pressure (measured in centimeters on 
the physiograph chart) while mentally imaging well-being 
(relaxing with pleasnt thoughts) was remarkably constant 
between self-esteem (se) groups: The high se was 2.99 cm 
(a = .88, N = 19), the middle se was 3.02 cm (a = .87, 
N = 32), and the low se was 2.96 cm (a = 1.08, N = 19). 

While resting comfortably the average systolic blood 
pressure of all subjects was 3.00 cm and no self-esteem 



-72- 



group mean deviated from the subject mean by more than .04 cm. 
Thus the "baseline" (relaxing) blood pressure for all three 
groups was virtually identical. 

When imaging depression the low self-esteem group's 
systolic blood pressure went up significantly (Duncan, DF = 
134, MS = 0.30, p < .05) more than the high self-esteem 
group (low se group mean increased 1.01 cm, high se in- 
creased 0.33 cm). The middle se group mean increased 0.75 cm 
and was significantly different from the other two (see 
Figure 4) . 

The systolic blood pressure that was first machine re- 
corded (first BP cuff inflation) was also significantly 
higher for those with low self-esteem compared to the high 
self-esteem group (low se increased 1.63 cm, high se in- 
creased .98 cm). The middle group mean increased 0.97 cm 
and was also significantly different from the low self-esteem 
group, but not from the high self-esteem group. This first 
BP cuff inflation was almost universally reported as a 
stressful event (98%). 

Cr ? : There will be no blood pressure differences 

between a younger group and an older age group 
during states of depression and stress (masking 
effect). 

There were no significant differences between the young- 
er and older age groups during the experimental treatments 
with systolic blood pressure as the dependent variable. 
Thus imaging well being (F < 1, p > .05), imaging dpression 
(F = 2.63, p > .05), and the event at the first BP cuff 
inflation (F = 2.73, p • .05), showed no systolic blood 



-73- 



CM mm Hg 

r 

L0 -H0 ^ 

3.0^100 ~ 
4.0 h 80 



-- D 
. W 



low middle high 
SELF-ESTEEM 



Figure 4. Systolic blood pressure during experimental 
treatments and its relationship to self- 
esteem. (Note: Middle se group has dis- 
proportionate number of older subjects.) 

First BP cuff inflation (stressful event) (S) 

Depression (D) 

Well being (W) 

Physiograph chart centimeters (CM) drawn to scale. 



-74- 



pressure differences between age groups. Thus the signifi- 
cant differences between high and low self-esteem, which 
raised the overall younger group mean, masked a priori blood 
pressure differences due to aging. 

Blood Pressure at Exit 

Blood pressure at exit (results of analyses of psychophysio- 
logical relationship between self-esteem and blood pressure 
measured independently of the physiograph and experimental 
treatments as a reward after subjects had successfully com- 
pleted all tasks) 

Systolic blood pressure at exit was not significantly 
correlated with the Tennessee Self-Concept Scale (r = -.17, 
p > .05), but diastolic was (r = -.33, p < .01). 

Subjects were again grouped by their score on the 
Tennessee Self-Concept Scale (TSCS) into high (upper 27%), 
middle (46%), and low (lower 27%) self-esteem groups. 

The exit systolic and diastolic blood pressures were 
analyzed using the analysis of variance procedure and Dun- 
can's post hoc statistic (see Tables 6 and 7). Subjects in 
the high self-esteem group had significantly lower blood 
pressure than did those in the low self-esteem group for 
both systolic (F(2,67) = 4.15, p < .02) and diastolic 
(F(2,67) = 5.50, p < .01) blood pressure (see Figure 5). 

The Duncan procedure confirmed that the high self- 
esteem group had significantly lower systolic blood pressure 
than did the low self-esteem group (DF = 67, MS = 167.17, 
p < .05). 



•75- 



Q 
W 
Prf 

D W 
CO H 

H 

W < 
Pi w 

p Pi 

CO H 
CO 

Pi<: 

Pm H 

o w 

OS 
O H 
,-) Pi 
PQ W 
Ph 

cj x 

M W 

J _ 

O S 
H O 

CO Pd 

CO 

>> 
o w 
Q<3 



o 

a 

CO o 

W Q 

W W 

Pi W 

O Pi 

W fe 
Q 



r^ r— I 






W 

Pi e 
p e 

CO ^ 

W >> 

Pi Pi 

Pm ^d 
U 

Q Pi 

O W 

O g 

►J 

« P-. 
O 
U 
h- ( 
hJ 

o 

H 



00 vO ON 
00 ro ro 



t— I .—I o 

I— I CN CN| 





a. 






3 




a. o 




3 


M 


ex 


o 


oo d 


m 







bO g 


^ 




<y 


bO 


H 


o 




<U 


j-j 


b 


0) 


05 


CJ 


u> 


(U 


0) 


w 


1 


iJ 


01 M-l 


o) 


1 


^ 


0.) 


iw 


CU 


i 


r-l 


CQ 


lh 


0) 




r^ 


02 


a) 


CD 




v— 1 


en 


x; i3 




b£)T3 


!2 


•H 


■H 


o 


SSJ 



-76- 



o 
w 
2 

p co 
co H 

<:s 
w w 
ss 

H 
w < 
Pd Pd 

5 3 

en H 

CO 

fa fa 

2 < 

Q fa 
OS 
O M 
fa Pd 
PQ W 

cu 

a x 
h w 
►J 

OS 
H O 
CO Pi 

< fa 



w 

S Pd 

<;<: 

S O 

CO 



fa 
o 

co o 
fa Q 
fa fa 
Pi fa 
O Pd 
fa fa 
Q 



O 
m 



^ oo 



en oo 

(Tsl LO 





a 

3 




o 




S-i 




M 




e 




CU 


w 


0) 


c ) 


■U 


Pi 


en 


L3 


0) 


o 


1 


CO 


CW 



25 60 

M X 
W g 

cd g 

|3 ^ 

CO 

CO >H 

fa Pd 
Pd p 

fa CJ 

Pd 

Q fa 

o 2 
o 

fa fa 

PQ O 

U 
H 

fa 
O 
H 
co 
< 



O CN O 



oo -<r vd 
vo r^ r-~ 





fa 






3 




a o 




3 


>~l 


a 


o 


60 3 


u 







60g 


u 




cu 


w 


B 


cu 




cu 


4J 


h 


OJ 


co 


0) 


-U 


a) 


QJ 


CO 


i 


4-J 


0) 


<4-t 


CO 


1 


.— 1 


QJ 


m 


cu 


i 


>— i 


CO 


4-1 


ai 




,— 1 


CO 


CD 


01 




.— 1 


CO 


,£ 13 




60 13 


\i 


•^ 


■H 


o 


fa S fa 



-77- 



mm Hg 



140 



120 



100 - 



80 



60 



Systolic 



Diastolic 



low middle high 
SELF-ESTEEM 



Figure 5. Systolic and diastolic blood pressure measured 
at exit and their relationship to self-esteem. 
(Note: Middle self-esteem group had dis- 
proportionate number of older subjects, and 
mean blood pressure for this group was elevated 
due to the aging process.) 



-71 



The middle group systolic blood pressure was not dif- 
ferent from the lower self-esteem group but was higher than 
the high self-esteem group. (Note: Middle group has dis- 
proportionate number of older subjects; Duncan, DF = 67, 
MS = 167.17, p < .05.) 

The Duncan procedure also confirmed that the high self- 
esteem group had significantly lower diastolic blood pres- 
sure than the low self-esteem group (DF = 67, MS = 58.84, 
p < .05). 

The middle self-esteem group diastolic blood pressure 
was not different from the lower self-esteem group but was 
higher than the high self-esteem group. (Note: Middle group 
has disproportionate number of older subjects; Duncan, DF = 
67, MS = 58.84, p < .05.) 

Systolic blood pressure was not significantly related 
to self-esteem for the younger age group (r = -.11, p > .05). 
However the 17 older subjects' SBP was highly related to 
their self-esteem (r = -.70, p < .01). 

The younger age group mean blood pressure was signifi- 
cantly different from the older group for systolic (F(l,68) = 
14.37, p < .001) blood pressure (see Appendix VI). The 
younger group systolic BP was 115.1 mm Hg (N = 53, a = 
12.0) and the older age group mean was 128.1 mm Hg (N = 17, 
o = 13.4). This exit difference confirmed an a priori dif- 
ference between age groups. This difference was found 
masked by the effect of self-esteem during the experi- 
mental treatments. 



79- 



There was no significant difference (Appendix VII) 
between the younger and older age group means for diastolic 
BP (F < 1, p > .05). The younger group mean was 73.6 mm Hg 
(N = 53, a = 8.1) and the older group mean was 72.5 mm Hg 
(N = 17, o = 7.8). Diastolic BP is much less reactive than 
systolic BP and remained high for the low se group composed 
mainly of younger people, even after more than one-half 
hour after machine data collection. Since there were sig- 
nificant differences between age groups at that time for 
systolic BP it is very reasonable to infer that, given enough 
time, there would be differences between age groups for 
diastolic blood pressure. Again, between group differ- 
ences were masked by the effect due to self-esteem. 

Reliability and Validity 

After completing the experimental treatments, selected 
subjects were given additional treatment (s) that were seem- 
ingly psychologically comparable to the original experimental 
treatments, a test-retest reliability. 

Fifty- three of the seventy subjects were requested to men- 
tally image guilt (a universally self -reported unpleasant 
image, as was imaging depression) and their psychophysiological 
measurements were compared with those these same subjects had 
displayed while they were imaging depression. Blood pressure, 
heart rate, and respiration rate were significantly cor- 
related between similar unpleasant experimental treatments 



80- 



(.71, .88, and .77, respectively; p < .0001). The electro- 
encephalograph (EEG) and galvanic skin response (GSR) were 
not correlated with their corresponding treatment (p > .05) 
(see Table 8) . 

Twenty- two subjects were read statements about death 
(reported as stressful by 91% of the subjects) and the 
psychophysiological measurements were correlated with the 
respective responses that took place at the first BP cuff 
inflation (also reported as stressful) . 

Three of the physiological responses were significantly 
correlated with their corresponding (stressful) experimental 
treatment responses: blood pressure at .57, p < .01; heart 
rate at .60, p < .01; and GSR at .68, p < .001. Two were 
suggestive that an N larger than twenty-two would be statis- 
tically significant: respiration rate at .36, p = .09; and 
EEG at .37, p = .08 (see Table 8). 

Further, correlations between the psychophysiological 
measurements and the TSCS from the two additional samples 
were compared to their respective psychologically similar 
event correlations with the expectation that the measures 
that had previously been found to be related to self-esteem 
would again be related. 

For the fifty-three subjects who additionally imaged 
guilt, the blood pressure and right forebrain electrical 
activity (EEG) were again related to self-esteem: r = -.38, 
p < .01 and r = -.34, p < .01, respectively. However, the 



1- 



H fa 

W 

iJ o 

W 

<J CM <-n 

HfQW 
Z H 

whs: 

S co w 

Pi > 

!Z H W 

W W 
W W 

rs sz d 

H W W 
WWW 
PP 3 co 
H W 
Cm W Pi 
H PQ H 

wp^ 

O <C co 
H H 

H -Z 

< --- W 
W CO g 
W H W 

oo pi 13 H 

w <: 

W Pi > H 

W M W CO 
PQ W 

< ffi H ffi 

H H S H 

O CO W 

<} W 

W O 

CO P-, 2 

H 2 M 

S p pi 

§ W 
W H ffi 
Pi W 
^ M O 
CO 13 2 

< O <] 

§ Q Z 
O 



-J 

u z < 

MOW 

o M w 

O CO 3 
W CO M 

o w 

M pi 
CO P-, 

>-i w 
X o 

Ph 

o 
ac 

a 

CO 

PL, 



o 

3 H 

W pi 

w <: 

3 W CO 

H PC H 

W !3 

pq W Q W 

W S § 

O U H 

M 2 <: 

H P-< O H 

< PQ M CO 

W H 

W H <d ffi 

p*i co W H 

pi pi w <; 

O M 2 W 

O W H Q 



Z 

W S 

W O H 

S M W 

H CO M 

W CO s 

PQ W O 

Pi 

Z Ph o 

O W 2 

H Q M 
H O 

< o < 

w z s 

W H H 

Pi O 

Pi 

O 



Q 
. . S3 

u m <; 



oo 

X> 



oo 
oo 



4-1 



a 






c 




0) 

t— i 






o 
a. 




CD 




01 


en 
a) 




C 




4-1 


^ 




•i-i 




CT3 




OJ 


cfl 




S-J 


C 


u 


^ 






•H 


3 


XI 




c 


4d 


7) 


0) 


01 


C 


W 


CO 


5-4 


4-1 


■H 




a 


o 


nj 


4J 


a 


M 


IW 


u 


c3 




p 


4J 


4-1 


■H 


nj 


T3 


x 


^1 


CL 


> 


C 


til! 


cfl 


'/J 


r—( 


'J 


•r-i 


oj 


1) 


(0 


^H 


Pi 


X 


Pi 


u 


XI 



-82- 



EEG for mentally imaging depression had not been signifi- 
cantly correlated with self-esteem but the well being state 
was significantly related (see Table 9). 

Heart rate, respiration rate, and galvanic skin response 
were again not significantly correlated (p > .05). Thus the 
five physiological measurements continued to act in approxi- 
mately the same manner as they had done earlier. 

Further, for the twenty- two subjects who were read 
statements about death, the psychophysiological responses 
were similar to their corresponding experimental treatment. 
Right forebrain activity was significantly correlated (r = 
-.41, p < .05) and blood pressure was suggestive that a 
larger sample would be statistically significant (r = -.39, 
p = .06). The other three measures were not significant 
(p > .05). Thus, the physiological responses had continued 
to act in approximately the same manner as they had in the 
original similar treatment (stressful event). 

Tables 3 through 7 show significant differences for 
each of the five psychophysiological measurements and their 
concurrent experimental treatment. Imaging well being, 
imaging depression, and the mental event at the first BP 
cuff inflation (stressful event) were shown to be signifi- 
cantly different from each other physiologically in five 
relatively independent ways (predictive validity) . 

In summary, the experimental treatments (independent 
variables) were significantly different psychological events 



p 




w 


Afl 


W 2; *-s 


PJW2 


p-t 


U W M 


en Q H 


S3 < 


en H O 


CJ wM 


CO ,-J 


HWHh 


wH W 


2 Dd 


SM« 


w s 


w H en 


H <J H 


en w Z 


WPdW 


1 H > 


Ch W 


hJ >" 


W PQ hJ 


co <; 


~H 


2 en 2 


WWW 
§ PQ 


H <d Pi 


W H < 
PQ Pi hJ 


<J H 


C^ CU > S 


H H 


WSHW 


t4W2 


PQ 2 W Jh 


< O Q hJ 


H H 2 hJ 


H W < 


<! P-> 


,JWH 


WOO 


2 ^0 


►J 


H O 


•• E5 X 


en w u 


H S >< 


►J w en 


PKh 


cn p 


w en 


&<z 


W i- 1 


2gO 


O P 


M _] _! 


H <C O 


hJ M M 


w 


PiO^ 


pi hJ en 


O O W 


UH,J 


en pq 


fe >" <C 


O X M 


Ph Pi 


>H O < 


Pi X > 


< 


S >" 


Is en 


5 P-. 



(jlnsas inassa"axs) 

siNawaivxs 

Hivaa dniwsh 



(iNaAa injssaaxs) 

NOIIViaNI 

aano da isdia 



(IN3A3 IMVSV31dm) 

nino 



(IN3A3 IMVSVSldNn) 
N0ISS3"dd3a 



ONI 39 TI3M 



2 



O 



1— 1 00 1— 1 cn o\ 

-johoh 



MD C\l <o 00 r~- 

CN OCNHtN 



nThhhoo 

CO CM 1— I r-l 00 









* 


o> 


r--. vo 


CTN OJ 











O CN 



HHCXHON 

CO O o o o 



CO 
CO 



O i-H 

r>« 1—1 

















QJ 


















QJ 


QJ 




H 




QJ 














QJ 5-1 


U 




3 




H 




ex 










CO 3 


3 




CO 




3 




3 






e 




C w 


CO 




CO 




CO 




a 






cfl 




O CO 


cn 




a 




CO 




3 u 






rH 




Ch QJ 


QJ 




h 




OJ 











be 




co h 


H 




Ph 




u 




M 











QJ QJ P-i 


Ph 








P-, 




O QJ 


MH 1 




1— 1 




4-> Pi 






13 








M 


O 4-1 




cti 




CO X) 


X) 




O 




T3 




QJ < 


r— 1 QJ 




43 




Pi c 







O 









M 


M JJr-l 




ex 




•H O 







rH 









< u 


C en rt 




Q) 




C ni rH 


rH 




pq 




rH 




QJ 







O 


a 


en pq 


pq 








pq 




}-i bO 


•h qj en 




c 


4-J 


•rH 




4-1 





4-J 




4-) 


QJ C 


4-1 OJ 


>> 


a) 


cn 


u 


-J 


•H 


•H 


•H 


a 


•rIT) 3 


CtS CO 4-1 


X3 


Pi 


ccj -H -H 


•H 


X 


rH 


X 


•H 


X rH O 


h to a 




M 




^ CH 


rH 


w 


O CJ 


rH 


W O r« 


qj qj a) 


QJ 


U) 


4-1 


•H Cfl O 


O 




4J 











^ C 


M 


U 


M 


a, > 4-j 


4J 


*J 


CO 


4-) 


4-J 


4-J 




^ C C 


O 


<U 


cc3 


COH CO 


CO 


CC3 


tfl 


V 


CO 


cd 




O dJ 


O 


r-i 


"J 


0) eg >, 


>, 




•H 




>, 






O H en 


W X pi en 


cn 




O 




CO 







•84- 



for each physiological measure and the measured physiology 
(dependent variables) acted in a moderately reliable manner 
during psychologically similar treatments. 



CHAPTER V 
DISCUSSION 



This study demonstrated that, for the subject popula- 
tion in this experiment, self-esteem was significantly re- 
lated to physiological indices (electroencephalogram and 
blood pressure) . During stressful events the effect seemed 
strong enough to exacerbate and did in fact mask a priori 
differences between younger and older age groups, a priori 
differences due to the aging process itself. Further, older 
subjects' systolic blood pressure was highly related to 
self-esteem even though they thought the experiment was over. 
This robustness suggests that the effect of self-esteem 
states on physiological indices has practical as well as 
theoretical implications. 

Dependent Variables 

Electroencephalogram 

Results of the analyses of the electroencephalogram 
(EEG) show that the EEG for right forebrain activity was 
directly related to self-esteem, especially during stressful 
events. During depression there was no relationship. How- 
ever, mentally imaging well being while relaxing was re- 
lated, as was guilt. 



•85- 



-86- 



The most surprising finding in the present experiment 
was that, when subjects were relaxing, right forebrain 
electrical activity was positively correlated with self- 
esteem. One possible explanation is the obvious, that those 
who have high self-esteem have more electrical activity 
because they engage in more active situational coping most 
of the time. 

Another possible explanation is that, since different 
self-esteem acts like different filters, the high and low 
self-esteem groups understood different things from the same 
instructions. For instance, the high self-esteem group may 
have heard a challenging task when the well being instruc- 
tions were read, while the low self-esteem group heard a 
chance to escape the testing situation at the same instruc- 
tions to relax and feel good. 

Neither of the above two psychological explanations are 
eliminated by a third explanation that is more physiologically 
based. During mentally imaging well being, people with low 
self-esteem, assuming they have suffered more psychological 
depression in the past, display less frontal activity than 
people with high self-esteem. This was suggested from Caciop- 
po and Petty (1981) who found that facial electromyograph 
displayed during mental imagery of the subjects' typical day 
was negatively related to the subjects' prior history of 
depression; and to Stanley et al. (1982) who showed depressed 
people are also likely to have lowered frontal activity 
because they have fewer frontal cortex imipramine binding 
sites . 



-87- 



The most probable explanation for a relationship be- 
tween self-esteem and EEG is a mix of all three psychological 
variables. High and low self-esteem groups often differ in 
coping strategies, hear different instructions, and have 
differing depression levels. 

Between group results showed the low self-esteem group 
had significantly more right forebrain activity than the 
high self-esteem group at the machine produced stressful 
event (first blood pressure cuff inflation) . This was repli- 
cated when a sample of the low self-esteem subjects again 
had more EEG activity when they were listening to state- 
ments about death. Further they had significantly more 
EEG activity while mentally imaging guilt (this also implies 
mentally imaging guilt and depression may be very different 
events: Guilt produced significantly more brain activity 
than did depression). 

Higher EEG activity for younger subjects appeared to be 
exacerbated by the effect of self-esteem during the machine 
produced stress. Again, this suggests that the effect of 
self-esteem states on right forebrain activity has practi- 
cal as well as theoretical implications. 

Electrocardiogram 

The statistical analysis used in the experiment showed 
that heart rate was not related to self-esteem. This lack 
of relationship could be a type II error (not finding an 



effect that is present) because the experimental manipula- 
tions were not successful in producing large enough changes 
in heart rate to produce significance. 

Heart rate was the most consistent intrasubject variable 
(imaging well being correlated .87 and .92, p < .01, with 
depression and stress, respectively). However, two con- 
founding variables were observed during data collection. 
Respiration rate affected heart rate as expected. But un- 
expectedly, many subjects showed a "between beat" varia- 
bility that is not reflected in the "beats per minute" 
statistic. For instance, one subject's heart rate looked like 
the following normal ECG:>L-X — / i~-^l- — \ — \~ — Aj — ~\] — \-\ • During 
the machine produced stress at the first blood pressure cuff 

inflation, this changed to: \\ J~\ H — \X -Wf"- 

These dramatically different (normal and arythmatic) beats 
are not different in beats per minute. 

Computer assisted technology ("on line" equipment) is 
recommended for subsequent research to quantify interbeat 
subleties as well as intrabeat variability for investigation 
involving personality variables and heart rate. 

Respiration Rate 

Respiration rate was not related to self-esteem in the 
present study. Again, type II error is suspected. 

During data collection several different "breathing 
styles" were observed. Respiration is sympathetic and 



•89- 



central nervous system activated: It is automatic but one 
can also breathe virtually any time one chooses. Simply 
counting data per minute was again not an adequate descrip- 
tion of the data. Some subjects breathe in quickly and let 
the air out slowly, others breathe in slowly and out quickly, 
Still others maintain a consistent breathing rate but change 
from shallow to deep breathing during stress. Rapid, 
shallow breathing under stress as reported in the litera- 
ture review was not observed in this study. This reaction 
was seen in the pilot studies immediately after the "startle 
response" and low self-esteem subjects, in particular, ex- 
hibited this "breathing style." (A "startle response" that 
was sometimes observed when the blood pressure cuff was 
machine inflated without specific and timely comment by 
the operator was seen as too dangerous to use with older 
subjects.) Again, more sophisticated analyses (computer 
assisted) looking at larger numbers of subjects may show a 
significant relationship between self-esteem and respiration 
rate . 

Perspiration Rate 

Perspiration rates were not significantly related to 
self-esteem in the present experiment. Some younger sub- 
jects had extremely high perspiration rates (the literature 
review showed this may be genetic). These high perspiration 
rate subjects, named "labiles," whose high rate is genetic, 



•90- 



will mask any effect due to self-esteem, especially among 
younger people. 

Blood Pressure 

Systolic blood pressure was found negatively related to 
self-esteem during periods of depression and stress during 
the experiment. After the experiment and debriefing (when 
blood pressure was taken at exit), casual systolic blood pres- 
sure was no longer correlated with self-esteem for the subject 
population. However, casual diastolic was negatively related 
(diastolic is less reactive than systolic). 

The casual systolic blood pressure at exit showed sig- 
nificant differences between the younger and older age groups 
and confirmed expected differences between age groups due to 
the aging process (see Appendix VI). However, this "a 
priori" difference was masked during the experimental mani- 
pulations due to the effect of self-esteem. 

Subjects with low self-esteem had suffered dispropor- 
tionately higher systolic blood pressure, blood pressure 
high enough to mask measured a priori differences between 
younger and older people. This speaks to the meaningful- 
ness of the relationship between self-esteem and blood 
pressure. 

The older age group's casual systolic blood pressure at 
exit was strongly related to self-esteem. This suggests self- 



■91- 



esteem plays an important part in the aging process it- 
self. 

Limitations of the Present Study 

The subject population was not random, but all indi- 
cations are that it was representative. The sample had 
approximately the same distribution on the Tennessee Self- 
Concept Scale as the norm group (see Appendix I). Further, 
the subjects' measured physiological characteristics gener- 
ally coincided with that expected from the literature 
review. 

The Tennessee Self -Concept Scale itself was a major 
limitation of the present study. It has the drawback of 
not being operationally "grounded," a limitation mentioned 
in the "Statement of the Problem" section of this study as 
being so serious as to call into question its validity. A 
subjective analysis of the content of the TSCS shows that 
more than one-half of the test items seem to reflect "guilt" 
("I ought to go to church more" and "I should trust my 
family more" are examples that seemingly reflect guilt) . 

The present study showed differing physiological 
indices for well being, depression, guilt, and stress. The 
weight of the physiological evidence presented in this study 
implies that the TSCS is a good beginning but is not measur- 
ing a self-esteem construct robust enough for optimal real- 
world prediction. 



•92- 



Correlations in this study of .20 to .70 predict about 
five to fifty percent of the variance. Although this is 
meaningful (i.e., enough to mask some effects of the aging 
process) this researcher expects future similar instruments 
to reflect much greater predictability, on the order of 
correlations around .85 to .90 and higher. Future instru- 
ments should entail a "match-up" (iteration) between physi- 
ology and factor analysis for subscales, each subscale cor- 
relating with self-esteem, but having little or no correla- 
tion between subscales. 

Iteration will involve culling items that seemingly 
reflect self-esteem but are not correlated with physio- 
logical indices and further culling dependent measures 
(physiological indices) that are not strongly correlated 
with self-esteem until an item pool is constructed that has 
validity and is grounded in physiology. 

The largest obstacle to the creation of such an instru- 
ment was also an obstacle in the present study. Equipment 
to measure the physiology itself has limitations such as 
inaccuracy and expense. However such obstacles are sur- 
mountable, albeit not easily. 

Certainly the limited number of subjects was a problem 
in the present study. This limited the number of variables 
that could be tested (e.g., self-esteem by age groups could 
not be calculated because of the small numbers of older 
people in the high and low self-esteem groups). However 
the "cup is half full" side of this issue is that effects 



■93- 



found with small numbers of subjects are most often 
robust. 

Suggestions for Additional Research 

Additional research empirically grounding self-esteem 
to physiology is warranted to operationalize theoretical 
constructs. Subscales concerned with the physiology of 
stress, depression, and guilt can be correlated with items 
from many of the currently available self -concept tests. 
Subscales with labels such as real-self and ideal self and 
other such labels already acceptable to many in the scien- 
tific community can be used to label the physiologically 
grounded constructs. 

Self-esteem research looking at Type A personality for 
shared physiological and psychological problems should be 
productive. Low self-esteem and Type A people often seem 
to share perf ectionistic (guilt?) and stressful traits. What 
are their similarities and differences and what does this 
imply for self-theory? Can the "essence" of the Type A per- 
sonality (with self-esteem and genetics factored out) become 
itself a subscale of a more comprehensive and predictive 
personality test? 

Further research is also warranted to investigate 
whether those with low self-esteem will improve physiologic- 
ally with rising self-esteem, as seems likely. "Real-world" 
investigation may soon be possible using advanced technology 



■94- 



(i.e., the Holter monitor records twenty- four hours of the 
electrocardiogram while the client goes about his everyday- 
routine) . 

Additional research investigating the role of self- 
esteem in the aging process is recommended. This study 
strongly suggests that having low self-esteem, which pro- 
duces concurrently elevated stress and depression levels, 
is a key factor in the aging process. The low self-esteem 
person seems to "practice" deleterious psychophysiology over 
his lifetime and accelerates the aging process. 

Implications for Education 

This study demonstrated that self-esteem is a meaning- 
ful construct with practical implications. Self-esteem was 
shown to be reflected in one's physiological characteristics 
in several ways. Which one "causes" the other is of little 
import to the teacher or clinician (e.g., a student with low 
self-esteem can learn relaxation, an adaptive coping strategy 
that reduces high blood pressure which in turn may help 
raise the student's self-esteem). 

However, the subtle interplay between cause and effect 
involving a "feedback loop," as described in the example 
above, is the stuff of self- theory. Untangling these nuances 
will enable construction of adequate scales for self-report 
instruments for the teacher and school psychologist. Thus 
armed the teacher can guide the student into the most 



•95- 



productive area for enhancement of the self. In addition, 
the teacher or counselor can make accurate prediction about 
students' "real-world" behavior and can then guide them into 
more productive learning situations and enhanced mental 
health. 

Implications for Theory 

Many theorists have constructed self-concept tests and/or 
posited a multiplex entity which is instrumental for mental 
health (e.g., Coopersmith, 1967; Fitts, 1972; and Hurlock, 
1974). 

Coopersmith (1967) tests self-constructs primarily 
along dimensions of confidence and anxiety. Fitts (1972), 
the author of the Tennessee Self -Concept Scale, tests thir- 
teen subscales and totals selected scores for an overall 
"Positive," or self-esteem score. Fitts' scales further 
cluster (using item analyses) to make what he labels empiri- 
cal scales: defensive positive, general maladjustment, 
psychosis, personality disorder, neurosis, personality inte- 
gration, number of deviant signs scores, number of integra- 
tive signs, and self -actualization scores. 

Fitts' (1972) self-esteem score includes what he calls 
identify, self-satisfaction, and behavior. He adds these 
scores to others for physical self, moral-ethical self, 
personal self, family self, and social self to get what he 
calls the Positive Score, a measure of self-esteem. 



•96- 



Hurlock (1974) seems to find a balance between Cooper- 
smith and Fitts. She posits four roles: the basic self, 
the ideal self, the transitory self, and the social self. 

Recently published results from longitudinal studies 
of longevity and life satisfaction suggest that the above 
theorists are on the right track. However, they may not be 
directly measuring the predictors found in the longitudinal 
studies. Palmore (1982), in his 25-year longitudinal study, 
found that the strongest independent predictors of longevity 
for men were health self -rating, work satisfaction, and per- 
formance intelligence. For women the best predictors were 
health satisfaction, past enjoyment of intercourse, and 
physical function rating. His predictors constituted a 
difference in longevity of 16 years for men and 23 years for 
women. Interestingly, "usefulness" was a minor predictor 
for men and not significant for women. 

"Self -perceived health," "not being neglected by 
friends," and "not riding in a car with a friend or neigh- 
bor as usual form of transportation" were significant pre- 
dictors of life satisfaction for the elderly in a three-year 
longitudinal study (Baur and Okun , 1983). 

Since the present study found that self-esteem accounts 
for as much as 50% of the effects in the aging process, 
future self-esteem instruments must reconcile the differ- 
ences between findings in the longitudinal studies and cur- 
rent theory. (In fact, because of the imperfect reliability 
and validity of the Tennessee Self -Concept Scale, the 



■97- 



correlations found in the present study are undoubtedly 
conservative and the effect on aging due to self-perception 
is probably higher than 50%.) 

Correlational investigations are always going to have 
certain kinds of problems with reliability and prediction. 
Inference is often made from behavior and self -report, but 
error-free inference, and hence prediction, may never be 
possible. Moreover, this researcher, given the human con- 
dition, certainly neither desires nor expects complete pre- 
diction. However, the technology is now present to improve 
prediction to a point where it can have meaningful positive 
future consequences for personality. For instance, in the 
longitudinal study above, "not riding in a car with a friend 
or neighbor as usual form of transportation" infers that 
"dependency on others" may be a predictor. Other research 
reported earlier in this study saying that "elderly men have 
a high rate of suicide" is additional inference that depen- 
dency may be a significant factor in self-esteem. "Past 
enjoyment of intercourse" suggests that longevity for women 
is related to the marriage relationship and thus the marriage 
relationship becomes a factor in self-esteem. Using infer- 
ence in the above ways to produce an item pool and then 
grounding the items in physiology should enhance prediction 
of longevity, success, and other "real-world" behavior. 

Construction of a self-concept instrument grounded in 
physiology would then make use of inferences from longitudinal 
studies, possibly items from the Coopersmith instrument and 



■98- 



the Tennessee Self -Concept Scale, and items generated by a 
theoretical model combining the above with Hur lock's (1974) 
model and the results of the present study (see Figure 6). 

The model suggested in Figure 6 deletes Hurlock's 
"transitory self" in favor of a more developmental approach. 
A theoretical essay suggesting this approach to the study of 
successful aging was presented by Ryff (1982) and her argu- 
ment seems compelling. She addresses issues of operational 
definitions, selective sampling, and stage theory. 

The theoretical model for construction of a self-con- 
cept instrument grounded in physiology is necessarily incom- 
plete at this point in time. More longitudinal data are 
needed. Further, additional research using the model itself 
will generate a more predictive model grounded in physiology. 
The projected model should have subscales that are not re- 
lated to each other, assuming factor analyses, but they 
should be related to the overall construct of self -concept . 
Then the educator and the counselor will have a tool that 
will enable successful and meaningful prediction. 

Conclusion 

"Self-esteem" can be operationalized using phsyiological 
variables such as right forebrain electrical activity and 
systolic blood pressure. This then may lead to operation- 
alization of more global constructs such as the self- 
concept . 



•99- 







Pd 






O 






M 






> 






jx! 






w 






pq 






H 






PL, 






W 






CJ 






2 






O 






U 






Pn 






hJ 






w 






CO 




►J 






< 






u 






M 


CO 




o w 




o u 




•J M 




o o 




M 


S 




CO 


H 




>H 






hj 






Pm 


CO 




!a 


M 




o 


C 




M 


•H 




H r-l 




O 


CU 




S 


0) 




W M-4 
















2^ 




O .-I 




M 


en 




H 


4-J 




H 


1 




2 M-l 




O .-i 




o 


0) 




o 


c/) 


CO 






pt 




CO 


.- 


5-1 


•H 


< 


o 


CO 


u 


•U 


> 


V. 


CJ i-H 


PC 


cti 


co 


;z> m-i 


a 


co 


s "' 


Cti 



o a) 


1 








t> 




c a t^> 


CO 












1« 4J-pl 


n 












E -h -h en 


cu 




CO 


CI) 






5-4 4-1 > CO CO 


cj 


a 


CO X 






o a ai cu co 


m 


o 


CU 


■H 






M-< -H bO 5-1 CU 


cti 


•H 


o 


a 






M-dChDC 


o 


4J 


o 


•H 






(U-d O M 


a 




2 


3 






D-, cti i— 1 cti 


•H 




CO 


CO 




1 










PL 


^\ 










o 


X ^ 






4-> 




O 4-J 


0) 






c 




<-* c 


N 






CU 




CU 


•H 




CO 


e 




^ B 


i-H U-l 




CJ 


c 




o a 


Cti i-l 




•H 


o 


U 


cti o 


M CU 




4-J 


5-i 


CU ,0 i-l 


0) CO 




CU 


•H 


4-1 X) CU 


C 




e 


>.-l 


a> > 


a) 




CU 


c 


■H 


cu cu 


bJD 




bO 0) 4-4 m-i X 



X CU 
O 5-i 



O O 



CO 



,4-1 CO t-H 
O (U^l 

cu co 

cu C cu X B 

ao) ^o 

co co cu ,n 



4-) 

n 

>^ cu 




QJ 
CJ 

CU 



cu 
o 

<U cti 

a fi 

cu cu o 
cti 



>. 



X <d CO X) CL-U 

■H cu u cu x a 

M-l CU H cti X) CU -H 

c3 CX CU hO^.-^.<4-l 

O f>, CO CU R C <4-l 

U H 3 M tI tI (U 



O 
•H 

4-1 

CO 

CJ 
•H 
M-l 

•H CO 
4-J CU 

S 3 

CU r-4 

X cti 
•H > 



C 

O o 
•H 
4-1 
CO 
4-1 

a 

<u 
ex 

cu 



H 4-J 

cti co >^ co 



CU -H CU 

5-1 e^ 

cd 4-i 

M-l O 



X 

cu^i^: 

> 4-1 4-) 

•H i-l bO 

cu co a 

CJ cu cu 

U Xi !-i 

CU 4-J 

CU CO 



CJ 

CO 
M-l 
CO 

XI -H 
4-1 4-J 
i-l CO 

cO co 

CU 

Xt 



iH 


CU 


CU 


CO 


CO 






i-l 


^H 


CO 


CO 


•H 


CU 


O 


X 


O 


M 


co 



cO <>• 
O M-l 
•H .-I 
CO cu 
!>.c/j 

Pm 



•H 

X 
CU 
X 

B 

o 
u 

M 
4-4 

c 
cu 

I 

3 



M-l 

o 



CJ 


5-4 

4-1 

CO 

C 
O 
CJ 

5-1 

O 
M-l 



CU 
X 
O 





cO >•, 
O bO 
•H O 

4-1 i-l 

CU o 
5-1 -H 
O CO 

cu >% 

rd.fi 



cu 

5-1 
=3 
bO 
•H 



-100- 



The findings in this study that depression, guilt, and 
well being states, in addition to stress, are related to 
self-esteem show that Cronbach and Snow (1977) and Corno et 
al. (1981), who were quoted in the literature review of this 
study, are in error when they characterized anxiety and 
self-esteem along a single dimension. This is so because 
the physiological effects found in this study show differing 
psychophysiology depending on the psychological event and the 
physiological variable and their relationships to self- 
esteem. This study demonstrates that self-esteem is an 
important, complex multidimensional concept grounded in one's 
physiology. 

The most important novel contribution, among the many 
significant results reported in this study, is that self- 
esteem is highly involved in the aging process itself. This 
speaks to the importance of longitudinal studies and per- 
sonality research, particularly self-esteem investigation. 



APPENDIX I 
SUBJECTS 



TSCS Norm Group Mean = 347.0 
a = 30.0. 

Study Subject Mean = 347.5; 
a = 30.4. 



SUBJECTS BY AGE GROUPS 

Age N = 70 



15-59 


N = 53 


60-78 


N = 17 


SUBJECTS BY SEX 




Sex 


N = 70 


Male 


N = 29 


Female 


N = 41 



There were no significant sex differences between the 
younger and older age groups (nonparametric Wilcoxon t-test 
approximation of 0.62 or p > .05). 

SEX BY SELF-ESTEEM 

There was no significant correlation between the sex 
of the subjects and self-esteem (r = .17, p > .05). 

SUBJECTS BY SOCIO-ECONOMIC STATUS (SES) 

SES N = 70 

Upper lower N = 35 
Lower middle N = 21 
Upper middle N = 14 

-101- 



102- 



There were no significant SES differences between the 
younger and older age groups (Wilcoxon _t-test approximation 
of 0.13 or p > .05). 

There were no subjects in the sample (N = 70) from the 
lower or the upper classifications of SES. 

SES BY SELF-ESTEEM 

There was no significant correlation between SES and 
self-esteem (r = 0.02, p > .05). 

SUBJECTS BY SES AND SEX 



se: 




SEX 


N 


= 


70 


Upper 


lower 


male 


N 


= 


13 






female 


N 


= 


22 


Lower 


middle 


male 


N 


= 


9 






female 


N 


= 


12 


Upper 


middle 


male 


N 


= 


7 






female 


N 


= 


7 



SES BY SEX 

Sex was approximately equally distributed within each 
socio-economic group. The non-parametric Wilcoxon 2- sample 
test gave a t_-test approximation significance of 0.45, or 
p > .05. Thus statistically, there was no significant dif- 
ference between the number of men and women in each socio- 
economic group. 

THERAPY BY SELF-ESTEEM 

There was no significant correlation between self-esteem 
and psychological therapy (r = -.06, p > .05). Further, 



•103- 



those who had therapy (N = 10) were approximately equally 
distributed among the self-esteem groups. The non-parametric 
Kruskal-Wallis test of chi-square approximation gave a chi- 
square of 1.61, p > .05, median one-way analysis. Thus 
statistically, there were not a significant number of those 
who had undergone therapy in any one self-esteem group. 



APPENDIX II 
INFORMED CONSENT 



You will be requested to sit down in an easy chair and relax 
while we monitor some of your body's physiology. 

With your help, three electrodes for an electrocardiogram 
and two for an encephalogram will be taped down by a re- 
searcher of your own sex. 

Your help will also be requested in placing the blood pres- 
sure cuff on your left arm and the finger perspiration 
monitor on two fingers of your right hand. 

After recording your responses while you relax for a couple 
of minutes we will read some statements from the self-concept 
forms you completed earlier. We want to see how your body 
physiology changes. We do not anticipate any discomfort or 
risk for you and will report your blood pressure to you when 
we turn the physiograph off. 

Please feel free to ask any questions now or at any time 
they arise, whether the physiograph is on or not. 

You are, of course, free to withdraw this consent at any time 
and discontinue without prejudice. 

No monetary compensation will be awarded. 



I understand if I am physically injured during this experi- 
ment, and if the experimenter is at fault, that 1 may seek 
appropriate compensation and may contact the Insurance Co- 
ordinator for information about compensation at 107 Tigert 
Hall, University of Florida, telephone number 392-1325. I 
understand that no other form of compensation is available. 

I have read and I understand the procedure described above. 
I agree to participate in the procedure and I have received 
a copy of this description. 



Subj ect 



Relationship if other than subject 



Witness 

George Michael Bedinger , 1403 Norman Hall, U. F, 
Principal investigator's name and address 



-105- 



APPENDIX III 

UNIVERSITY OF FLORIDA HUMAN SUBJECTS 

COMMITTEE APPROVAL 



UNIVERSITY OF FLORIDA 

INSTITUTIONAL REVIEW BOARD 

114 PSYCHOLOGY BUILDING 

GAINESVILLE, FL 32611 

1904) - 392 - 0433 



May 1, 1981 



TO: Mr. George Michael Bedinger 
1403 NRN 

FROM: C. Michael Levy, Chair<§^5-' 

University of Florida Institutional 
Review Board 

SUBJECT: Approval of University of Florida Institutional 
Review Board Project # 81-73 
"Physiological Correlates of Self-Esteem" 

Your request for approval for a research project involving human 
subjects, referenced above, is approved as recommended by the 
University of Florida Institutional Review Board. The Board 
has concluded that subjects are placed at risk in the approved 
research. It is essential that you obtain written, witnessed 
informed consent (including the insurance statement) from each 
participant. You are reminded that a change in protocol in this 
project must be approved by re-submission of the project to the 
Board for approval. 

If the project has not been completed by April 28, 1982, 
please telephone our office (2-0433) and request instructions 
for obtaining a renewal of this approval. 

By a copy of this memorandum, your department Chair is reminded 
that she or he is responsible for being informed about the stat- 
us of all projects involving human subjects in your department, 
and for reviewing the protocol of such projects as often as 
necessary to insure that each project is being conducted in the 
manner approved by this memorandum. 



CML/her 



Dr. Donald Avila 

Dr. R. R. Sherman 

Dr. D. C. Smith 

Dr. F. M. Wahl 



1403 NRN 

314 NRN 

140 NRN 

(unfunded) 237 GRI 



•107- 



APPENDIX IV 
MEDICAL HISTORY AND SES QUESTIONNAIRE 



CONFIDENTIAL 



Do you have any important medical history that may affect your recordings, such 
m epilepsy, high blood pressure, cardiac disease, or emphysema? Are you in . 
counseling now in either individual or group therapy? When did you start? 
Have you completed counseling and, if so, when? 

(Please use back if necessary) 



Have you taken any drugs or other medication today? 

What is your occupation? 

What is your education? 

What is your father's occupation? 

Do you have any questions? 

Thank you. 

! 2 3 4 5 6 7 8 9 10 11 12 13 14 15 15 17 18 

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 

35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 

(All information will be kept confidential) 



109- 



APPENDIX V 
INSTRUCTIONS TO SUBJECTS FOR MENTALLY 
IMAGING TREATMENT CONDITIONS 



WELL BEING 

Initial instructions: 

"I would like for you now to mentally image feeling 
good. Please relax as much as possible and think of taking 
a trip to a place that you really like. 

Go someplace like the beach or to the mountains. Go 
with someone or by yourself. I want you to really enjoy where 
you are while you are relaxing. Try to really experience 
where you are. For instance if you go to the beach I want 
you to smell the salt air and feel the ocean breeze. Feel 
the hot sun on your skin. Relaxing and enjoying. ... If 
you go to the mountains I want you to feel the cool mountain 
air, hear the wind in the trees. Relaxing and enjoying. 
Feeling so good. ..." 

During data collection: 
As blood pressure cuff inflates: "relaxing and enjoying. . ." 
after deflation: "feeling so good. ..." 

DEPRESSION 

Initial instructions : 

"I would like for you now to mentally image feeling bad. 
Please think of something bad that has happened to you in 



110- 



Ill 



the past. Something like not having enough money or a close 
loved one lying in a casket. Try to really experience where 
you are. For instance if you think of being out of money 
try to recall the feeling you had when you saw the bills 
and wondered what you were going to do. Feeling so bad. . . 
If you think of a lost loved one I want you to see the cas- 
ket closing. . . . You'll never see this loved one again. 
Feeling so bad. ..." 

During data collection: 
As blood pressure cuff inflates : "what am I going to do? . 
after deflation: "feeling so bad. ..." 

GUILT 

Initial instructions: 

"I would like for you now to mentally image feeling 
guilty. Please think of something you did in the past that 
you really wish you had not done. Or think of a time when 
you said something to someone that you later wished you had 
not said. Try to really relive the guilty feelings and say 
things to yourself like 'I really wish I hadn't done that' 
or 'I wish I could take back what I said. '" 

During data collection: 
As blood pressure cuff inflates: "I really wish I hadn't 
done that. . . ."; after deflation: "I really wish I could 
do that over. ..." 



112- 



STRESS (instruction induced) 

Initial instructions: 

"I'm now going to read you five statements about death. 
I want you to think of yourself as I read them." 

During data collection: 
As blood pressure cuff inflates: "I am not at all afraid 
to die. . . ."; after deflation: "I am not at all afraid 
to die. . . ." 



APPENDIX VI 
RESULTS OF ANALYSIS OF VARIANCE PERFORMED ON SYSTOLIC 
BLOOD PRESSURE MEASURED SEPARATELY FROM PHYSIOGRAPH 
AND ITS RELATIONSHIP BETWEEN YOUNGER AND 
OLDER AGE GROUPS 



3 

w 

S 

PL. CO 

W Ph 

00 1=5 

O 

o cm 
w o 
pd 
p w 

CO o 

<< 
w 

W Q 

Pi hJ 
P o 

CO 
CO Q 

w z 

Pd <d 

Ph 

Pd 

Q W 
O O 
O S 

►J P 
PQ O 

O 
M Z 

hJ W 
O W 
HIS 
co H 
>< w 
co pq 

2 P-. 
O H 

Q co 
PJ 2 
S O 
PS H 

O H 
Ph < 
Pd XI 
PJ CrJ 
Pl, pi 

W CO 
O H 
13 H 

M Q 
Ph Z 
< < 
> 

fe Ph 

°S 

CO O 
H O 
CO M 
>h CO 

►J >i 

<: x 

Z Ph 

Ph O 
O Ph" 
Cm 
CO 
H 

CO 
PJ 

Pi 



w 

S Pd 

<; <; 
w p 

CO 



o 

s 

CO O 
W Q 
W PJ 
Pi W 

O Cm 
PJ Ph 

p 



3 




o 




M 




00 


^ 




o 


0> 


u 


W) 


u 


< 


PJ 



2 



— 1 


CO 


PL, 


— 1 


T— 1 


3 
O 
Sh 
bO 

(U 
50 

c 

0) 
01 

•u 


M 


to 


0> 




X 


X) 


R 


s 


4-J 


id 


e 


0) 


— 1 


.— i 


!h 
0) 



in 



Ph 






P0 






O 


!h 




Pi 


O 




U 


GO 


i-i 




c 


J 


PJ 


3 


T3 


u 


o 


— 1 


< 


>- 


o 



!h 
QJ 
■P 

01 

6 

•H 
.-I 
r-1 
•H 



>^ 



cfl 




O 




•H 




4H 




■H 




C 




tu: 


•-* 


•H 




co 


HO 




X 


CO 




•rH 


R 




R 


0J 




M T3 


p 


0) 


CO 


j-j 


CO 


en 


0) 


•r-l 


lH 


> 


CU OJ 




u 


T3 XI 


O XI 


O 


cfl 


.— 1 




X 


CO 




•H 


U 




•r-( 


>, 


^H 


M 





3 



• 11. 



APPENDIX VII 

RESULTS OF ANALYSIS OF VARIANCE PERFORMED ON DIASTOLIC 

BLOOD PRESSURE MEASURED SEPARATELY FROM PHYSIOGRAPH 

AND ITS RELATIONSHIP BETWEEN YOUNGER AND 

OLDER AGE GROUPS 



Pm 



< < 

fa C3 

en 



fa 
o 

S 
en o 
W Q 
w w 
Pi w 
O Pi 
fa fa 
Q 



i— I vO 





ex 






p 






o 




fa 


5-1 




t 3 


M 


u 


P^ 




o 


ED 


01 


u 


O 


M 


U 


en 


< 


fa 



bO 



bO 



fa 






ID 






O 


u 




ai 


CU 




o 


00 


u 




C 


a 


fa 


3 


X) 


CO 


o 


r-H 


< 


>H 


o 



fa 

O 

bC 

(D 
bO 
cd 

C 

cu 
cu 

-U 
0) 
fa 



c 




ctf 




u 




•H 




•rl 




(3 




bi: 




•H 




CO 






*-S 


-U 




O 


toll 


C fa 


to 


e 


cfl 


)-> 


r* 






X) 


0) 


CJ 


u 


u 


d 


ct) 


CO 


■H 


co 


> 


CD 


a) 


M 


H 


fafa 




fa 


T3 


cd 


O 







CO 


fa 


•H 


fa 






>> 


a 


U 


■H 


3 


r— I 


o 


o 


u 



co S 



116- 



BIBLIOGRAPHY 



Appel, M. , Gorkin, L., & Holroyd, K. Cardiovascular 

response to interpersonal response. Psychophysiology , 
1981, 18 (2), 162. ~ 

Avila, D.L. Psychophysiological study of self-concept theory. 
Unpublished research proposal, University of Florida, 
1980, p. 2. 

Ax, A.F. The physiological differentiation between fear and 
anger in humans. Psychosomatic Medicine , 1953, 15 , 
433_442. " 

Bandura, A. Self-ef f icacy mechanism in human agency. 
American Psychologist , 1982, 3_7 (2), 122-147. 

Baur, P. A., and Okum, M.A. Stability of life satisfaction 

in late life. The Gerontologist , 1983, 23 (3), 261-265. 

Beck, A.T. An inventory for measuring depression. 
Archives of General Psychiatry , 1961, 4, 561. 

Bedinger, M. , Bedinger, S., & Purkey, W. Project Self- 

Discovery: Psychology and Myself . University of Florida 
Division of Continuing Education, in press. 

Benson, H. The Relaxation Response . New York: William 
Morrow, 1975. 

Bettelheim, B. Individual and mass behavior in extreme 

situations. Journal of Abnormal and Social Psychology , 
1943, 38, 417-452. 

Bibring, E. The mechanisms of depression. In Greenacre, P. 
(Ed.), Affective Disorders . New York: International 
Univer sities Press, 1953 . 

Brazier, M.A.B. The analysis of brain waves. In Teyler , 

T.J. (Ed.), Readings from Scientific American: Altered 
States of Awareness ! San Francisco: W.H. Freeman and 
Company, 1962. 

Breslau, L. , & Haug , M. Depression and Aging . New York: 
Springer Publishing Company , 1983 . 



117- 



■118- 



Bultena, J. Ten year longitudinal study of the aged. 

Journal of Gerontology , 1978 (Sep.), 33 (5), 748-754. 

Burish, T.G. Type A/B behavior patterns. In Woody, R.H. 
(Ed.), Encyclopedia of Clinical Assessment , Volume I. 
San Francisco: Jossey-Bass Publishers, 1981. 

Burns, D. Feeling Good: The New Mood Therapy . New York: 
Morrow, 1980. 

Cacioppo, J.T., & Petty, R.E. Electromyograms as measures 
of extent and affectivity of information processing. 
American Psychologist , 1981 (5), 441-456. 

Cammer , L. Up from Depression . New York: Simon and 
Schuster, 1972. 

Campbell, K.B., Baribeau-Braun , J., & Braun, C. Neuroana- 
tomical and physiological foundations of extraversion. 
Psychophysiology , 1981, 18 (3), 263-272. 

Cannon, W.B. The mechanism of emotional disturbance of 

bodily functions. The New England Journal of Medicine , 
1928, 198. (17), 877^B7T 

Clynes, M. Biocybernetics of space-time forms in the gene- 
sis and communication of emotion. In symposium, "Bio- 
cybernetics of the dynamic communication of emotions 
and qualities," presented at the meeting of the American 
Association of Science, Chicago, 1970. 

Combs, A.W. , Avila, D.L. , & Purkey , W.W. Helping Relation - 
ships: Basic Concepts for the Helping Professions 
(2nd ed.). Boston: Allyn and Bacon, 1978. 

Combs, A.W. , Richards, A.C., & Richards, F. Perceptual 
Psychology: A Humanistic Approach to the Study of 
Persons (3rd ed. ) . New York: Harper and Row, 1976. 

Combs, A.W. , & Snygg , D. Individual Behavior: A Perceptual 

Approach to Behavior . New York: Harper & Brother, 1959. 

Coopersmith, S. The Antecedents of Self-Esteem . San Fran- 
cisco: Freeman, 1967. 

Corno, L. , Mitman, A., & Hedges, C. The influence of direct 
instruction on student self -appraisals : A hierarchical 
analysis of treatment and aptitude- treatment interaction 
effects. American Educational Research Journal , 1981, 
18 (1) , 39-61. 

Cronbach, L.J., & Snow, R.E. Aptitudes and Instructional 

Methods: A Handbook for Research on Interactions . New 
York: Irvington/Naiburg , 1977. 



-119- 



Davison, L. , & Valins, S. Cognition and false feedback. 
Journal of Personality and Social Psychology , 1972, 
3, 242-251. 

Dienstbier, R.A. Emotion- attribution theory: Establishing 

roots and exploring future perspectives. In Murray, S., 
and Levine, R. (Eds.), Nebraska Symposium on Motivation , 
Volume 26. Lincoln, Nebraska: University of Nebraska 
Press, 1979. 

Dunbar, H.F. Psychosomatic Diagnosis . New York: Hoeber, 1943 

Fabry, J.J. Depression. In Woody, R.H. (Ed.), Encyclopedia 
of Clinical Assessment , Volume II. San Francisco: 
Jossey-Bass Publishers, 1981, p. 588-601. 

Fenz, W.D., & Epstein, S. Gradients of physiological arousal 
in parachutists as a function of an approaching jump. 
Psychosomatic Medicine , 1967, 2^9, 33-51. 

Fitts, W.H. The Self Concept and Performance . Nashville, 
Tennessee: Counselor Recordings and Tests, 1972. 

Friedman, M. , & Rosenman, R.H. Overt behavior patterns in 
coronary disease. Journal of the American Medical 
Association , 1960, 1_73, 1320-1325. 

Gaitz, CM. Depression in the elderly. In Fann , W.E., and 
others (Eds.), Phenomenology and Treatment of Depres - 
sion . New York: Spectrum, 1977. 

Gale, A. The psychophysiology of individual differences: 
Studies of extraversion and the EEG. In Kline, P. 
(Ed.), New Approaches in Psychological Measurement . 
New York: John Wiley and Sons, 1973, p. 215. 

Garwood, M. , & Engel, B.T. Age differences in individual 
specificity. Psychophysiology , 1981, 18^ (2), 139. 

Gatchel, R.J. Learned helplessness (physiological concomi- 
tants). In Woody, R.H. (Ed.), Encyclopedia of Clinical 
Assessment , Volume I. San Francisco: Jossey-Bass 
Publishers, 1981. 

Gersten, J., Langer , T. , Eisenberg, J., & Orzeck, L. Child 
behavior and life events: Undesirable change or change 
per se? In Dohrenwend, B.S., and Dohrenwend, B.P. 
(Eds.), Stressful Life Events: Their Nature and Effects . 
New York: John Wiley and Sons, 1974. 

Glass, D.C. Behavior Patterns, Stress, and Coronary Disease . 
New York! Wiley, 1977. 



120- 



Graham, D.T. Cutaneous vascular reactions in Raynaud's 

disease in states of hostility, anxiety and depression. 
Psychosomatic Medicine , 1955, 17 (3), 200-217. 

Graham, D.T. , Stern, J. A., & Winokur , G. Experimental in- 
vestigation of the specificity of attitude hypothesis 
in psychosomatic disease. Psychosomatic Medicine , 
1958, 20 (6), 446-457. 

Hamachek, D.E. Encounters With the Self (2nd ed. ) . New 
York: Holt, Rinehart and Winston, 1978. 

Hastrup, T.L. Effects of electrodermal lability and intro- 
version on vigilance decrement. Psychophysiology , 
1979, 16 (3), 302-309. 

Hilgard, E.T., Atkinson, R.L. , & Atkinson, R.C. Introduction 
to Psychology . New York: Harcourt Brace Jovanovich, 
Inc., 1979. 

Howsam, R.B., Corrigan, E.C., Denemark, G.W., & Nash, J. 
Educating a Profession . Washington, D . C . : American 
Association of Colleges for Teacher Education, 1976. 

Hurlock, E.B. Personality Development . New York: McGraw- 
Hill Inc. , 1974. 

Hypertension in Adults . (DHHS Publication PHS 81-1671.) 
Hyattsville, Maryland: National Center for Health 
Statistics, U.S. Department of Health and Human Ser- 
vices, 1981. 

Kemple, C. Rorschach method and psychosomatic diagnosis: 

Personality traits of patients with rheumatic disease, 
hypertensive cardiovascular disease, coronary occlusion 
and fracture. Psychosomatic Medicine , 1945, 7_, 85-89. 

Krech, D., Crutchfield, R. , & Livson, N. Elements of 
Psychology . New York: Alfred A. Knopf, 1974. 

Lacey, T.I., Kagan, J., Lacey , B.C., & Moss, H.A. The viscer- 
al level : Situational determinants and behavioral 
correlates of autonomic response patterns. In Knapp , 
P.H. (Ed.), Expression of the Emotions in Man . New 
York: International University Press, 1963. 

Lacey, T.I., & Lacey, B.C. Some autonomic-central nervous sys- 
tem interrelationships. In Black, P. (Ed.), Physiological 
Correlates of Emotion . New York: Academic Press, 1970. 

Lang, P.J., Rice, D.G., & Sternbach, R.A. The psychophysi- 
ology of emotion. In Greenfield, H.S., and Sternbach, 
R.A. (Eds.), Handbook of Psychophysiology . New York: 
Holt, Rinehart and Winston, 1973. 



121- 



Lazarus, R.S. A cognitive analysis of biofeedback control. 
In Schwartz, G.E., and Beatty, J. (Eds.), Biofeedback : 
Theory and Research . New York: Academic Press, 1977, 
p. 85. 

Lazarus, R.S., Cohen, T.B., Folkman, S., Kanner , A., & 
Schaefer, C. Psychological stress and adaptation: 
Some unresolved issues. In Selye, H. (Ed.), Guide to 
Stress Research . New York: Van No strand Reinhold, 
1980. 

Levenson, R.W. , & Ditto, W.B. Individual differences in 
ability to control heart rate: A test of the impor- 
tance of personality, physiological and other variables. 
Psychophysiology , 1981, 18_ (1), 154. 

Light, K.C. Cardiovascular responses to effortful active 

coping: Implications of stress in hypertension develop- 
ment. Psychophysiologist , 1981, L8 (3), 216-224. 

Marinelli, R.P. Anxiety. In Woody, R.H. (Ed.), Encyclopedia 
of Clinical Assessment , Volume I. San Francisco: 
Jossey-Bass Publishers, 1981. 

Mason, J.W. Emotion as reflected in patterns of endocrine 
regulation. In Levi. L. (Ed.), Emotions: Their 
Parameters and Measurement . New York: Raven Press, 
I975\ 

Mason, J.W., Maker, J.T., Hartley, L.H., Morigey, E., Per- 

low, M. J. , & Jones, L.G. Selectivity of corticosteroid 
and catecholamine response to various natural stimuli. 
In Serban, G. (Ed.), Psychopathology of Human Adapta - 
tion . New York: Plenum, 1976. 

McConnell, J.V. Understanding Human Behavior: An Intro - 
duction to Psychology (3rd ed. ) . New York: Holt, 
Rinehart and Winston, 1980. 

Menninger , K.A. , & Menninger, W.C. Psychoanalytic observa- 
tions in cardiac disorders. American Heart Journal , 
1936, U, 10-21. 

Miller, R.G., Rubin, R.T., Clark, B.R., Crawford, W.R., & 

Arthur, R.J. The stress of aircraft carrier landings: 
I. Cortico-steroid responses in naval aviators. 
Psychosomatic Medicine , 1970, 32_, 581-588. 

Mischel, W. Introduction to Personality (2nd ed.). New 
York: Holt, Rinehart and Winston, 1976. 

Mittelmann, B., & Wolff, H.G. Affective states and skin 

temperature: Experimental study of subjects with "Cold 
Hands" and Raynaud's Syndrome. Psychosomatic Medicine , 
1939, 1 (2), 271-293. 



122- 



Mortality , Volume II. Hyattsville, Maryland: National Center 
for Health Statistics, U.S. Department of Health and 
Human Services, 1979. 

Mulder, G. , & Mulder, L.J.M. Information processing and 
cardiovascular control. Psychophysiology , 1981, 1_8 
(4) , 392-401. 

NARCO Biosystems Physiograph Manual . Houston, Texas: NARCO 
Biosystems, Inc., 1980. 

Nisbett, R.E., & Schachter, S. Cognitive manipulation of pain. 
Journal of Experimental Social Psychology , 1966, 2, 227-236 

Osier, W. Lectures on Angina Pectoris and Allied States . 
New York: Appleton-Century-Crof ts , 1897. 

Palmore, E.B. Predictors of the longevity difference: A 25- 
year follow-up. The Gerontologist , 1982, 2_2 (6), 513-518. 

Richter, C.P. On the phenomenon of sudden death in animals 
and man. Psychosomatic Medicine , 1957, 1_9, 191-198. 

Rosier, F. Event-related brain potentials in a stimulus- 
discrimination learning paradigm. Psychophysiology , 
1981, 18 (4), 447-461. 

Ryff, CD. Successful aging: A developmental approach. The 
Gerontologist , 1982, 11 (2), 209-214. 

Schachter, S., & Singer, J.E. Cognitive, social, and physio- 
logical determinants of emotional state. Psychological 
Review , 1962, 69, 379-399. 

Schaffer, L.F. Fear and courage in aerial combat. Journal 
of Counseling Psychology , 1947, 11 , 137-143. 

Schandry, R. Heart beat perception and emotional experience. 
Psychophysiology , 1981, 19 (4), 475-488. 

Schell, A.M., & Lusche, D.J. Psychophysiological response 
patterns of coronary prone and non-coronary prone 
males. Psychophysiology , 1981, 18 (2), 139. 

Schildkraut, J., & Freyhan , F.A. Neuropharmacological studies 
of mood disorder. In Zubin, J. (Ed.), Disorders of 
Mood . New York: Grune & Stratton, 197~2T 

Schildkraut, J., & Kety, S.S. Biogenic amines and emotions. 
Science , 1967, 156, 21-30. 

Schucker, B., & Jacobs, D.R. Assessment of behavioral risk 
for coronary disease by voice characteristics. Psycho - 
somatic Medicine, 1977, 39, 219-228. 



123- 



Schuyler, D. The Depressive Spectrum . New York : Aronson, 1974. 

Schwartz, G. , & Beatty, J. Biofeedback: Theory and Research . 
New York: Academic Press, 1977. 

Schwartz, G. , Flair, P., Man del, M. , Salt, P., Mieske, M. , 
& Klerman, G. Facial electromyography in the assess- 
ment of improvement in depression. Psychosomatic 
Medicine , 1978, 40 (4), 355-360. ~~ ' 

Seligman, M.E.P. Helplessness . San Francisco: Freeman, 1975. 

Spence , K.W. Types of constructs in psychology. In Marx, 
M.H. (Ed.), Theories in Contemporary Psychology . New 
York: Macmillan, 1963, p. 170. 

Spielberger, C.D. Anxiety: State-trait-process. In Spiel- 
berger , C.D., and Sarason, I.G. (Eds.), Stress and Anxi - 
ety , Volume I. Washington, D.C.: Hemisphere, 1975. 

Stanley, M. Tritiated imipramine binding sites are decreased 
in the frontal cortex of suicides. Science , 1982, 216 
(18 June), 1337-1340. 

Stern, J. A., Winskur , G. , Graham, D.T., & Graham, F.K. Al- 
terations in physiological measures during experimentally 
induced attitudes. Journal of Psychosomatic Research , 
1961, 5, 73-82. 

Sternbach, R.A. Principles of Psychophysiology . New York: 
Academic Press, 1966. 

Terman, L.M. , & Oden , M.H. The Gifted Group at Mid-Life . 
Stanford: Stanford University Press, 1959. 

Valins, S. Cognitive effects of false heart-rate feedback. 
Journal of Personality and Social Psychology , 1966, 4_, 
400-408. 

Van Egeren, L.F., Abelson, J.L., & Thornton, D.W. Cardio- 
vascular consequences of expressing anger in a mutually 
dependent relationship. Journal of Psychosomatic 
Research , 1978, 22, 537-548. 

Weg, R.B. Changing physiology of aging: Normal and patho- 
logical. In Woodruff, D.S., and Birren, J.E. (Eds.), 
Aging . New York: D. Van Nostrand Company, 1975. 

Weller, G.M., & Bell, R.O. Basal skin conductance and neo- 
natal state. Child Development , 1965, 36, 647-657. 

Wilier, J.C. Anticipation of pain-produced stress. Electro- 
physiological study in man. Physiology and Behavior , 
1980, 25 (1), 49-51. 



-124- 



Wolf, S., & Wolff, H.G. Human Gastric Function . New York: 



Oxford University Press, 1947 



Wortman, C, & Lof tus , E. 
Knopf, 1981. 



Psychology . New York: Alfred A. 



Wylie, R.C. The Self -Concept : A Review of Methodological 
Considerations and Measuring Instruments (rev. ed. ) , 
Volume I. Lincoln, Nebraska: University of Nebraska 
Press, 1974, p. 316. (a) 



Wylie, R.C. The Self-Concept: Theory and Research on 
Selected Topics (rev. ed.), Volume II. Lincoln, 
Nebraska: University of Nebraska Press, 1974. (b) 



BIOGRAPHICAL SKETCH 

George Michael Bedinger was born May 8, 1939, in 
Charleston, WV. He graduated from Stonewall Jackson High 
School in Charleston and began college at Hampden- Sydney 
College, VA, in 1957. 

After two years he transferred to Marshall University 
where he continued a concentration in chemistry and zoology 
and a major in psychology. For the four summers between 
sessions he was employed in the chemical laboratories of 
Union Carbide Corporation. 

In early 1962 he was ordered to active duty in the U.S. 
Army and served five years . Upon returning from Vietnam he 
resigned his commission to continue his formal education. 

Bedinger graduated from the University of Charleston 
in 1968 with a B.A. in psychology. He graduated from the 
University of South Florida in 1974 with an M.A. in poli- 
tical science and a community college teaching certificate. 

He taught for six years at Tampa College where he also 
served at various times in the following capacities: chair- 
man of the social sciences department, dean of the St. 
Petersburg campus, and assistant dean of the college. 

He was an adjunct instructor at Hillsborough Community 
College and is currently an adjunct instructor at Lake 



125- 



-126- 



City Community College and Central Florida Community 
College in addition to employment at the University of 
Florida (UF) . 

While at UF in his doctoral program in educational 
psychology, Bedinger was employed as a graduate teaching 
assistant and also earned a Certificate in Gerontology. 
At UF he also presented a poster session (as second author) 
in the Experimental Psychology division of the American 
Psychological Association. 

Bedinger' s current employment at UF is in the Division 
of Continuing Education where he has rewritten (first 
author) and instructs a high school psychology course 
(workbook format) which is implemented in many Florida 
high schools. He also is presently serving on the State 
of Florida Life Skills Task Force (Division of Community 
Colleges and Department of Corrections) and has authored 
articles in their publications. 



I certify that I have read this study and that in my 
opinion it conforms to acceptable standards of scholarly 
presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



t^u^--iXu 



Donald L. Avila, Chairman 
Professor, Foundations of 
Education 



I certify that I have read this study and that in my 
opinion it conforms to acceptable standards of scholarly 
presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



/I 



VL 



i_L 



6- A_ 



Barry Guinagh 

Associate Professor, Foundations 

of Education 



I certify that I have read this study and that in my 
opinion it conforms to acceptable standards of scholarly 
presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 




Robert Jester ^ ' 

Associate Professor, Foundations 

of Education 



I certify that I have read this study and that in my 
opinion it conforms to acceptable standards of scholarly 
presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



-Wv-c,! fa Lv 



Harold Riker 

Professor, Counselor Education 



This dissertation was submitted to the Graduate Faculty of 
the Department of Foundations of Education in the College 
of Education and to the Graduate Council, and was accepted 
as partial fulfillment of the requirements for the degree 
of Doctor of Philosophy. 

August 1983 



^V Ax \\LvA. ^ 



Acting Chairman, Foundations of 
Education 



Dean for Graduate Studies and 
Research 



UNWEBSlTYOFF^'j, 



f\SS 08285 231 9